- Email: [email protected]
The Relationship between Objective Frequency–Volume Chart Data and the I-PSS in Men with Lower Urinary Tract Symptoms
european urology 52 (2007) 811–818
available at www.sciencedirect.com journal homepage: www.europeanurology.com
Benign Prostatic Enlargement – Bladder Outlet Obstruction
The Relationship between Objective Frequency–Volume Chart Data and the I-PSS in Men with Lower Urinary Tract Symptoms Tet L. Yap a,b,*, David A. Cromwell a, Christian Brown a, Jan van der Meulen a, Mark Emberton a,b a b
Clinical Effectiveness Unit, Royal College of Surgeons of England, London, UK Institute of Urology and Nephrology, University College Hospital, London, UK
Article info
Abstract
Article history: Accepted January 4, 2007 Published online ahead of print on January 12, 2007
Objective: The International Prostate Symptom Score (I-PSS) is considered a benchmark in defining the severity of lower urinary tract symptoms (LUTS). However, its relationship with variables of the frequency–volume (FV) chart, an objective measure of LUTS, is not well understood. We analyzed this relationship between the I-PSS and FV chart variables in men with LUTS. Methods: Patients referred for the investigation of uncomplicated LUTS completed an I-PSS and a 3-d FV chart. Correlation coefficients were used to describe the relationship between total I-PSS and its components with the FV variables. The distributions of FV variable scores within the categories of I-PSS severity were also examined. The effect of patient’s quality of life score on I-PSS and FV variables was analyzed using regression. Results: A total of 104 patients completed both the I-PSS and FV chart. The association between FV variables and I-PSS scores was generally weak, with 24-h frequency being the strongest (r = 0.43) and with considerable overlap within each severity category. There was weak association between the I-PSS for nocturia and urgency and their FV counterparts. The number of selfreported nocturnal voids tended to be overestimated. These discrepancies were partly explained by the patient’s quality of life rating, which reduced the level of variation between I-PSS score and FV variables and accounted for some of the overestimation. Conclusion: Weak associations were found between the self-rated I-PSS scores and FV measures of voiding, suggesting that the accurate assessment of LUTS requires self-reported measures of symptoms and objective measures of voiding behaviour. Significant differences between the two types of data might be explained by perceived quality of life.
Keywords: AUA Quality of Life Score Benign prostatic hyperplasia Frequency–volume charts International Prostate Symptom Score Lower urinary tract symptoms
# 2007 European Association of Urology. Published by Elsevier B.V. All rights reserved. * Corresponding author. Clinical Effectiveness Unit, Royal College of Surgeons of England, 35–43 Lincoln’s Inn Fields, London WC2A 3PE, UK. Tel. +44 077 806 11631; Fax: +44 020 7869 6644. E-mail address: [email protected] (T.L. Yap). 0302-2838/$ – see back matter # 2007 European Association of Urology. Published by Elsevier B.V. All rights reserved.
doi:10.1016/j.eururo.2007.01.013
812
1.
european urology 52 (2007) 811–818
Introduction
Lower urinary tract symptoms (LUTS) affect more than 40% of men over 60 yr of age in the United States and Europe [1,2]. The majority of these symptoms are attributed to benign prostatic hyperplasia (BPH), but they can also arise from other causes [3]. It is now common to distinguish between three types of LUTS: storage symptoms, such as frequency and urgency; voiding symptoms, such as slow stream; and postmicturition symptoms, such as incomplete emptying [4]. The International Prostate Symptom Score (I-PSS), a questionnaire that is widely used to assess LUTS in patients, covers all three types, measuring incomplete emptying, frequency, intermittency, urgency, weak stream, hesitancy, and nocturia. A final question (the AUA Quality of Life Score, AUAQoL) asks about how the symptoms affect the patient’s quality of life. A patient gives each symptom a rating from 0 to 5 based on its severity over the past month. These are summed to produce an overall severity score, which is typically used to categorize patients as having mild (score 0–7), moderate (8–19), or severe (20–35) symptoms [5]. However, there has been increasing debate on the precision and accuracy of the I-PSS. First, patient recall may introduce errors. Second, voiding behaviour varies naturally over time, and about one in five patients might be expected to have their total I-PSS score fluctuate by an average of five points, when measured within 1 mo [6]. This can lead to changes in the categorization of LUTS severity in more than one-quarter of men with BPH [7]. Third, symptom questionnaires have been shown to significantly overestimate the severity of symptoms like nocturia and frequency [8,9]. The implications of these findings are that treatment decisions and conclusions about treatment effectiveness based on the I-PSS are more arbitrary than they might appear [7]. A different measure of voiding behaviour is the frequency–volume (FV) chart. This noninvasive instrument is likely to be less susceptible to recall bias and allows the assessment of LUTS outside the artificial setting of a clinic. These charts have been used since the 1970s to investigate LUTS [10], as well as to monitor the outcomes of therapeutic interventions such as bladder training [11]. The precision of chart-recorded data in describing storage symptoms like nocturia, frequency, and urgency have prompted many to recommend its wider use in clinical practice [5,8,11,12]. Despite their widespread use, the relationship between the different FV chart variables and the
I-PSS is not well understood. For instance, it has been noted that frequency measured using an FV chart is consistency lower than patients’ responses to a symptom questionnaire, and it has been suggested that patients who are more disturbed by their symptoms may tend to overestimate their condition when answering a question [11]. The distribution of values for FV variables within I-PSS severity groups (mild, moderate, severe) have yet to be examined. Also, the relationship between individual I-PSS components and related FV variables remains unclear. In this study, we investigated the relationship between the I-PSS and FV chart variables in men with LUTS. We analyzed the association between FV variables and the total I-PSS score, as well as the association between selected FV variables and individual I-PSS questions. Finally, we examined whether the quality of life question of the I-PSS can explain any differences in the severity of symptoms measured using the I-PSS and the FV chart. 2.
Methods
Between February 2003 and June 2004, 140 men with uncomplicated LUTS were recruited from the outpatient department of two urological centres in London. All patients over 40 yr of age who were referred for the first time by their general practitioner to one of the two participating urological outpatient departments with LUTS were eligible for inclusion. Men were excluded: (1) if they had received any form of medical therapy (a-blocker, 5-a reductase inhibitor, or anticholinergic) or had prostatic surgery or pelvic radiotherapy within the prior 3 mo; (2) if they had severe symptoms necessitating immediate medical or surgical treatment; (3) if they had complications potentially related to their symptoms (PSA over 4 ng ml 1, residual volumes over 200 ml, creatinine above 130 mmol l 1, bladder stones, haematuria, urinary retention, recurrent urinary tract infections); (4) if they were unable to speak or understand English; or (5) if they had uncontrolled diabetes, dementia, or end-stage cardiac or respiratory failure. Patients completed an I-PSS questionnaire during their outpatient visit after they had been enrolled in the study and were given a 3-d FV chart to complete within 14 d. On the chart, patients recorded their waking time and time to bed for 3 d, along with volume of liquids drunk to the closest hour and volume of urine voided to the closest hour (measured using a calibrated jug). The parameters extracted from the FV charts were symptom-related measures, such as 24-h frequency, nocturnal frequency, daytime frequency, and number of urgent episodes, and volume-related measures, such as 24-h total intake, 24-h total voided volume, mean voided volume, maximum voided volume, and nocturnal volume. All were calculated as defined by the International Continence Society [4], with frequency and mean volume variables being expressed per 24 h.
813
0.17 0.05 0.04 0.22 0.11 0.17 0.21 0.38* 0.12 0.36* 0.23 0.28* 0.17 0.27* 0.17 0.20 0.24 0.15 0.36* 0.28* 0.28* 0.18 0.18 0.11 0.13 0.11 0.04 0.20 0.18 0.13 0.17 0.20 0.11 0.21 0.08 0.10
0.05 0.05 0.04 0.01 0.08 0.24 0.07 0.03 0.11
0.02 0.02 0.02 0.06 0.08 0.24 0.08 0.06 0.11
0.41* 0.28* 0.22 0.20 0.15 0.12 0.15 0.23 0.04
0.22 0.06 0.04 0.08 0.04 0.09 0.12 0.43* 0.06
Nocturnal volume per 24 h Episodes of urgency per 24 h Episodes of nocturia per 24 h Daytime frequency 24-h frequency Maximum voided volume Mean voided volume Total voided volume per 24 h
Correlation (either g or r) is significant at 0.01 level.
All 140 patients completed the I-PSS questionnaire; 104 men completed at least 1 d of the FV chart, and
*
Results
Total I-PSS Incomplete emptying (Q1) Frequency (Q2) Intermittency (Q3) Urgency (Q4) Weak stream (Q5) Straining (Q6) Nocturia (Q7) AUA-QoL score
3.
Symptom
The general characteristics of the enrolled patients were described using summary statistics (ie, mean and SD). The direct relationship between the total I-PSS score and the FV variables was analyzed using Pearson correlation. The association between the quality of life rating and the FV variables was analyzed using Spearman rank correlation due to the short scale of the AUA-QoL. Analysis of variance was then used to determine whether there were systematic differences between distributions of the FV variables and the categories of I-PSS severity. The total I-PSS score was categorized in the standard way, using 0–7 to denote mild symptoms, 8–19 to denote moderate symptoms, and 20–35 to denote severe symptoms [13]. Box plots were used to show the distribution of values for those variables whose mean differed across the I-PSS categories. Because the total I-PSS score was the sum of the individual question scores, multiple linear regression was used to assess whether the total I-PSS score was related to a combination of FV variables. Stepwise regression was used to find the most parsimonious model. The significance level for removing variables was set at p = 0.2 [14]. The relationship between individual I-PSS questions and FV variables was analyzed using Spearman rank correlation and the gamma statistic (g). The gamma statistic is preferable to rank correlation if the data contain many tied observations because both variables are measured on short scales [14]. Like other measures of association, the gamma statistic ranges from 1 to 1 and is 0 when there is no association. The analysis focussed on those FV variables expected to be strongly related to particular I-PSS questions. The 24-h frequency was compared to the I-PSS questions on incomplete emptying and frequency. The number of urgent voids was compared to the I-PSS questions on urgency and frequency. Finally, the FV variables for the nocturnal frequency and nocturnal volume were compared to the I-PSS question on nocturia. Both instruments define nocturia as the number of nocturnal voids. This enabled us to examine whether selfreported values were overestimated. Multiple regression was used to assess the degree of underreporting and whether this might be explained by perceived quality of life as measured using the AUA-QoL. All other I-PSS questions ask patients to rate symptom burden as: 0 for ‘‘not at all,’’ 1 for ‘‘less than one in five times,’’ 2 for ‘‘less than half the time,’’ 3 for ‘‘about half the time,’’ 4 for ‘‘more than half the time,’’ and 5 for ‘‘almost always.’’ This scale was not equivalent to the scales used by the FV variables, and this difference could weaken the association between similar questions. In the case of urgency, this possibility could be examined by transforming the variable into an equivalent scale. First, the percentage of the total voids that were urgent was calculated. The transformed variable was then defined based on increasing 20% point intervals (0% was assigned a score of 0, 1–20% was assigned a score of 1, and so on).
Total intake volume per 24 h
Statistical analysis
Table 1 – Relationship between individual frequency–volume variables and I-PSS
2.1.
0.04 0.09 0.18 0.03 0.19 0.20 0.01 0.15 0.13
european urology 52 (2007) 811–818
814
european urology 52 (2007) 811–818
Table 2 – Mean and standard deviation of frequency–volume chart variables by I-PSS severity category (mild, 0–7; moderate, 8–19; severe, 20–35) Frequency–volume chart variable 24-h frequency Daytime frequency Nocturnal frequency Number of urgent episodes Total voided volume per 24 h Mean voided volume Maximum voided volume Nocturnal volume per 24 h Total intake volume per 24 h
Mild I-PSS (n = 6) mean (SD)
Moderate I-PSS (n = 76) mean (SD)
Severe I-PSS (n = 22) mean (SD)
9.1 (3.5) 7.6 (3.4) 1.5 (1.1) 1.2 (2.3) 1573 (736) 172 (32) 408 (66) 510 (195) 1990 (1110)
9.1 (2.2) 7.4 (2.0) 1.7 (0.9) 2.1 (2.6) 1707 (671) 188 (62) 372 (150) 584 (277) 1891 (750)
11.0 (3.3) 8.7 (2.8) 2.2 (1.3) 3.8 (3.7) 1848 (723) 172 (62) 376 (177) 683 (333) 1918 (602)
96 patients completed all 3 d. Among the men completing both instruments, the mean age was 63 yr (SD = 11, range: 40–86) and the mean symptom duration was 4.4 yr (SD = 6.2). Most men were Caucasian (81 of 104, 78%) and had some academic qualification (75 of 104, 72%). The mean scores on the I-PSS and AUA-QoL were 16.2 (SD = 6.1) and 3.6 (SD = 1.1), respectively. There was no important difference in age, ethnicity, level of education, mean symptom duration, I-PSS symptom, or AUA-QoL scores between the 104 men who supplied frequency–volume data and the 36 men who did not. The associations between total I-PSS score and the FV variables were highest for the variables derived from the number of voids (Table 1). The 24-h frequency (r = 0.41) and number of urgent episodes (r = 0.36) showed the largest (although only moderate) correlations. Daytime frequency and total I-PSS showed a similar correlation (r = 0.36), but the daytime values were strongly correlated with 24-h frequency (r = 0.95). There was no association between total I-PSS and the variables based on volume voided. There was also no association between AUA-QoL scores and any of the FV variables. The mean values of the different FV variables did not exhibit a strong trend across the categories of I-PSS severity (Table 2). Indeed, consistent increases were observed for only four variables, and the differences between severity groups were only statistically significant for the mean number of voids per 24 h and episodes of urgency. For both these variables, however, the distribution of FV values within each I-PSS severity category overlapped substantially (Figs. 1, 2), demonstrating that the I-PSS severity categories do not discriminate effectively between patients with different voiding behaviours. In the regression model developed to assess whether total I-PSS could be predicted from a combination of all FV chart variables, only the coefficients for 24-h frequency and urgency were
ANOVA F-test p value 0.01 0.06 0.06 0.04 0.59 0.50 0.85 0.26 0.95
statistically significant at a 5% level, being 0.76 (95% CI: 0.33–1.12) and 0.55 (95% CI: 0.16–0.95), respectively. The model explained only 22% of the variation in total I-PSS.
Fig. 1 – Distribution of 24-h frequency from frequency– volume chart data by I-PSS severity category (mild, 0–7; moderate, 8–19; severe, 20–35).
Fig. 2 – Distribution of urgent voids from frequency–volume chart data by I-PSS severity category (mild, 0–7; moderate, 8–19; severe, 20–35).
815
european urology 52 (2007) 811–818
Table 3 – Mean and standard deviation of frequency–volume chart-recorded variables aggregated by scores on related I-PSS questions FV chart variable
I-PSS question
I-PSS question score 0
24-h frequency
1
2
Correlation
3
4
5
Q1 (incomplete emptying) Q2 (frequency)
8.9 (2.0)
8.6 (2.2)
9.3 (2.4)
10.4 (3.3)
10.3 (2.6)
12.9 (2.3)
r = 0.25
9.3 (0.5)
7.6 (2.2)
8.9 (2.0)
9.5 (2.9)
9.8 (1.9)
10.9 (3.6)
r = 0.33*
Number of urgent episodes
Q2 (frequency) Q4 (urgency) Q4 (urgency) transformed
2.8 (4.0) 1.9 (2.9) 1.3 (1.7)
1.3 (1.5) 1.5 (2.0) 1.1 (1.2)
1.8 (3.2) 2.0 (2.8) 1.3 (1.4)
1.9 (2.4) 2.1 (2.5) 1.4 (1.5)
2.8 (2.5) 1.9 (2.3) 1.4 (1.4)
4.1 (4.2) 5.5 (3.8) 2.9 (1.4)
r = 0.29* r = 0.29* g = 0.27
Nocturnal frequency Nocturnal volume (ml)
Q7 (nocturia) Q7 (nocturia)
0.6 (0.1) 513 (53)
1.2 (0.7) 465 (238)
1.6 (0.8) 538 (165)
2.0 (0.9) 645 (326)
2.3 (0.8) 718 (269)
3.2 (1.9) 917 (452)
r = 0.44* r = 0.32*
*
Correlations are significant at 0.01 level.
Table 3 summarizes the association between the individual I-PSS questions and the FV variables. Most FV variables did not show any significant relationship with their I-PSS counterparts. The association between I-PSS question 4 (urgency) and the equivalent FV variable was weak (r = 0.29). This did not appear to be due to the different scales, because the association was also weak between the I-PSS question and the transformed urgent FV variable (g = 0.27). Indeed, the effect of the transformation was small and the normal and transformed urgent FV variables were highly correlated (Spearman’s r = 0.97). There was a positive association between nocturnal frequency and the I-PSS nocturia question, although the agreement was not strong (g = 0.43). The self-reported nocturnal voids tended to be higher than those actually recorded on the FV chart (Fig. 3), with the discrepancy increasing as the number of self-reported voids increased. Patients who reported no nocturnal voiding on the I-PSS were
Fig. 3 – Distribution of nocturnal frequency from frequency–volume chart data across the scores on I-PSS question 7 (nocturia).
the exception, with the majority recording some nocturnal voids on their FV chart. Regression was used to examine whether the difference between the I-PSS and FV-chart nocturnal frequency was related to perceived quality of life due to a patient’s urinary symptoms. The predicted I-PSS was estimated to increase by only 0.5 voids for every unit increase in FV nocturnal frequency (95% CI: 0.3–0.7). The predicted I-PSS was estimated to increase by 0.3 for every unit increase in the AUA-QoL variable, suggesting differences in quality of life ratings explains some of the overestimation. Nonetheless, the regression model could only explain 31% of the variation between patients in I-PSS nocturnal voids.
4.
Discussion
This study showed that the total I-PSS correlated moderately with FV chart symptom measures of frequency and urgency but not with the FV volume variables. When patients were categorized according to their I-PSS severity, the mean severity of frequency and urgency FV variables increased across the mild, moderate, and severe groups. However, there was a large overlap in the distributions of FV values between these groups, and knowing a patient’s individual value would not allow a clinician to accurately infer what their severity category might be. The poor association between individual variables and the total I-PSS did not seem to be due to the aggregate construction of the total I-PSS. When we assessed how well a combination of all FV variables could predict the total I-PSS, only 24-h frequency and urgency were found to explain some of the observed variation in total I-PSS. The correlation between individual FV variables and related I-PSS questions was also found to range
816
european urology 52 (2007) 811–818
from weak to moderate. The different scales used by the I-PSS questions and the FV variables may be a source of variation that limited the correlation; in the case of the two measures of urgency, however, transforming the FV variable to an equivalent scale did not improve the correspondence between the measurements. The higher scores on I-PSS question for nocturia significantly overestimated chart recordings, which was consistent with other studies of patients referred to urological clinics [8,9,15]. It has been hypothesized that patients who are more disturbed by their symptoms may tend to overestimate their condition [11]. Our results seem to support this. The AUA-QoL score gives an indication of how urinary symptoms affect a patient’s quality of life and was found to partially explain some of the overestimation. The I-PSS correlates poorly with certain objective measures of urinary symptoms, such as mean and maximum flow rates, prostate size, and postvoid residual volume [16]. Few other studies have examined the relationship between FV chart data and the I-PSS in patients with LUTS. Ku et al. reported poor agreement between total I-PSS and FV chart variables in 164 patients referred for LUTS evaluation, noting that many overestimated their nocturia on I-PSS [17]. Another study of 160 men with LUTS found some correlation between small voided volumes (Kendall and Gibbons coefficient, r = 0.17), nocturia (r = 0.21), and daytime frequency (r = 0.31) with total I-PSS [15]. That study used only 24-h FV charts, however, which have been shown to produce less reliable values due to the amount of day-to-day variation in voiding behaviour [5,12]. The lack of correlation between the two instruments seen with our data may be due to a number of study limitations. First, only 74% of patients completed at least 1 d of the FV charts. Nevertheless, this compliance rate is similar to rates reported by other studies using FV charts [9,17]. Moreover, there were no differences in the characteristics of patients, their total I-PSS, or the AUA-QoL score between responders and nonresponders ( p > 0.5 for all). Second, the weak correlation could be due to poor chart recording. To reduce this risk, all subjects were taught to fill the charts by a specialist nurse. Analysis was performed using data from patients who completed charts for 1, 2, and 3 d to ensure the analyzed sample was as representative as possible. In fact, it appeared that this group of patients did not differ significantly from patients who completed charts for all 3 d, and repeating the analysis using only data from 3-d charts produced equivalent results. Third, the group of men enrolled in the study may be more homogenous than patient
populations seen normally in clinics. This would cause the correlation coefficients to be underestimated. However, the range of scores observed in our sample was similar to those reported elsewhere [18,19] and does not seem to be a substantial source of bias. Fourth, the patients completed the FV chart after being seen in the clinic, whereas patients’ I-PSS answers refer to their symptoms over the past month. Consequently, adjacent periods in time are being compared and it is possible that the poor agreement was due to changes in patients’ underlying conditions. This seems unlikely, however, as LUTS is a chronic condition, and the patient cohort did not report any adverse consequences such as infections or acute retention during their clinical assessment. Finally, although no actual medical or surgical intervention was performed, the enrolled patients saw a medical professional in the assessment clinic. This may have influenced their perception of their symptoms and consequently the prospectively collected FV data. However, there is no evidence to suggest that a single visit would result in such large changes. The differences in the information provided by the two measures of LUTS suggest that the two are complementary assessment tools. The Fifth International Consultation on BPH recommended the use of both the I-PSS (with the AUA-QoL question) and FV charts (especially in patients with nocturia) in initial LUTS evaluation [20]. The lack of association between the I-PSS and frequency measures suggests that the FV chart is better suited to providing information regarding the aetiology of symptoms. For instance, mean voided volume can be reduced by a decrease in functional bladder capacity (eg, due to overactive bladder disorders). Also, FV volume data such as nocturnal volume are important in distinguishing causes of nocturia, including nocturnal polyuria [21]. Another reason to use both tools is the apparent difference in self-reported and actual voiding behaviour. The I-PSS for nocturia overestimated chart recordings, and using both tools allows for objective and perceived levels of symptom severity to be differentiated, both when a patient is assessed and when monitoring the effect of treatment. Our results suggest that the severity of nocturia reported by patients is influenced by the way patients perceive their quality of life due to their urinary symptoms, and this may also have an influence on their health-care seeking behaviour. A trial using a population-based sample showed an underestimation of nocturia compared to FV measurements [8]. The way in which men’s experiences and perceptions may influence their self-reported symptoms is
european urology 52 (2007) 811–818
an area that requires further investigation. The AUA-QoL is a single question embodying a specific interpretation of quality of life, namely, the perceived burden of urinary symptoms. However, quality of life is a complex concept that covers bother, physical and mental health, functional impairment, and social integration. Bother reflects the overall distress of having LUTS and is strongly related to social anxiety. Furthermore, LUTS affects patient’s self-perception, leading to restrictions in social activities [22]. Validated instruments exist to measure these dimensions of quality of life, including condition-specific measures such as the BPH Impact Index, the DAN-PSS, and the ICSQoL [23,24]. We suggest that the apparent overestimation of selfreported symptoms could be better explained if data from these measures could be incorporated into an analysis of the relationship between the I-PSS and FV chart variables.
Conflicts of interest We acknowledge (1) that all authors have made a substantial contribution to the information or material submitted for publication; (2) that all have read and approved the final manuscript; (3) that we have no direct or indirect commercial financial incentive associated with publishing the article.
References [1] Speakman MJ, Kirby RS, Joyce A, Abrams P, Pocock R. Guideline for the primary care management of male lower urinary tract symptoms. BJU Int 2004;93:985–90. [2] Platz EA, Smit E, Curhan GC, Nyberg LM, Giovannucci E. Prevalence of and racial/ethnic variation in lower urinary tract symptoms and noncancer prostate surgery in U.S. men. Urology 2002;59:877–83. [3] Jones D. Benign prostatic hypertrophy and lower urinary tract dysfunction. In: Weiss RM, George NRG, O’Reilly PH, editors. Comprehensive urology. London: Mosby International; 2001. p. 452. [4] Abrams P, Cardozo L, Fall M, et al. The standardisation of terminology of lower urinary tract function: report from the Standardisation Sub-committee of the International Continence Society. Neurourol Urodyn 2002;21:167–78. [5] Palnaes HC, Klarskov P. The accuracy of the frequency– volume chart: comparison of self-reported and measured volumes. Br J Urol 1998;81:709–11. [6] Barry MJ, Girman CJ, O’Leary MP, et al., Benign Prostatic Hyperplasia Treatment Outcomes Study Group. Using repeated measures of symptom score, uroflowmetry and prostate specific antigen in the clinical management of prostate disease. J Urol 1995;153:99–103.
817
[7] O’Connor RC, Bales GT, Avila D, Gerber GS. Variability of the International Prostate Symptom Score in men with lower urinary tract symptoms. Scand J Urol Nephrol 2003;37:35–7. [8] McCormack M, Infante-Rivard C, Schick E. Agreement between clinical methods of measurement of urinary frequency and functional bladder capacity. Br J Urol 1992;69:17–21. [9] Blanker MH, Bohnen AM, Groeneveld FP, Bernsen RM, Prins A, Ruud Bosch JL. Normal voiding patterns and determinants of increased diurnal and nocturnal voiding frequency in elderly men. J Urol 2000;164:1201–5. [10] Torrens MJ. The effect of selective sacral nerve blocks on vesical and urethral function. J Urol 1974;112:204–5. [11] Abrams P, Klevmark B. Frequency volume charts: an indispensable part of lower urinary tract assessment. Scand J Urol Nephrol Suppl 1996;179:47–53. [12] Bryan NP, Chapple CR. Frequency volume charts in the assessment and evaluation of treatment: How should we use them? Eur Urol 2004;46:636–40. [13] Barry MJ, Fowler Jr FJ, O’Leary MP, et al. The Measurement Committee of the American Urological Association. The American Urological Association symptom index for benign prostatic hyperplasia. J Urol 1992;148: 1549–57. [14] Siegel S, Castellan Jr NJ. Measures of association and their tests of significance. In: Siegel S, Castellan Jr NJ, editors. Nonparametric statistics for the behavioural sciences. New York: McGraw Hill; 1988. p. 291–3. [15] Van Venrooij G-EPM, Eckhardt MD, Gisolf K-WH, Boon TA. Data from frequency–volume charts versus symptom scores and quality of life score in men with lower urinary tract symptoms due to benign prostatic hyperplasia. Eur Urol 2001;39:42–7. [16] Wadie BS, Ibrahim EH, de la Rosette JJ, Gomha MA, Ghoneim MA. The relationship of the International Prostate Symptom Score and objective parameters for diagnosing bladder outlet obstruction. Part I: when statistics fail. J Urol 2001;165:32–4. [17] Ku JH, Hong SK, Kim HH, Paick JS, Lee SE, Oh SJ. Is questionnaire enough to assess number of nocturic episodes? Prospective comparative study between data from questionnaire and frequency–volume charts. Urology 2004;64:966–9. [18] Bertaccini A, Vassallo F, Martino F, et al. Symptoms, bothersomeness and quality of life in patients with LUTS suggestive of BPH. Eur Urol 2001;40(Suppl 1):13–8. [19] Ezz ED, Kiemeney LA, de Wildt MJ, Debruyne FM, de la Rosette JJ. Correlation between uroflowmetry, prostate volume, postvoid residue, and lower urinary tract symptoms as measured by the International Prostate Symptom Score. Urology 1996;48:393–7. [20] Roehrborn CG. Focus on lower urinary tract symptoms: nomenclature, diagnosis, and treatment options—highlights from the 5th International Consultation on Benign Prostatic Hyperplasia, June 25–27, 2000, Paris, France. Rev Urol 2001;3:139–45. [21] Weiss JP. Nocturia: ‘‘do the math’’. J Urol 2006;175: S16–8.
818
european urology 52 (2007) 811–818
[22] Glover L, Gannon K, McLoughlin J, Emberton M. Men’s experiences of having lower urinary tract symptoms: factors relating to bother. BJU Int 2004;94:563–7. [23] Donovan JL, Kay HE, Peters TJ, et al. Using the ICSQoL to measure the impact of lower urinary tract symptoms on quality of life: evidence from the International
Editorial Comment on: The Relationship between Objective Frequency–Volume Chart Data and the I-PSS in Men with Lower Urinary Tract Symptoms Cosimo De Nunzio, Department of Urology, Ospedale Sant’Andrea, Roma, Italy [email protected] According to the 5th International Consultation on BPH [1], recommended tests to be performed in all patients with lower urinary tract symptoms (LUTS) and benign prostatic enlargement (BPE) included medical history, quantification of symptoms and bother, physical examination, urinalysis, prostate-specific antigen, and frequency–volume (FV) chart. FV charts are simple to complete and provide useful clinical information as an estimation of minimum voided volume, mean voided volume, maximum voided volume, number of voids in 24 h, mean diuria, mean voided volume in daytime, mean nocturia, mean voided volume at night; and mean fluid intake in daytime and at night. In patients with nocturia, FV charts are considered the most valid instrument to distinguish between nocturia, nocturnal polyuria, or excessive fluid intake [2]. Data from FV charts has been used to compare a patient’s voiding habits at home and the urodynamic variables assessed by uroflowmetry and filling cystometry [3]. The authors of this report [4] evaluated the relationship between FV charts data and the severity of symptoms and bother assessed by the International Prostatic Symptom Score (I-PSS) in men with LUTS. Overall a weak association was detected between total I-PSS and FV variables with the 24-h frequency (r = 0.41) and the number of urgent episodes (r = 0.36) showing the largest but moderate correlations. No associations between FV variables and quality-of-life score were observed. Most FV variables did not show any significant relationship with their I-PSS counterparts and a positive but not strong association was revealed between nocturnal frequency and the I-PSS nocturia question with the highest scores on the I-PSS question for nocturia, overestimating chart
Continence Society–Benign Prostatic Hyperplasia Study. Br J Urol 1997;80:712–21. [24] Haltbakk J, Hanestad BR, Hunskaar S. Use and misuse of the concept of quality of life in evaluating surgical treatments for lower urinary tract symptoms. BJU Int 2003; 91:380–8.
recordings. The lack of correlation between these two measures of LUTS is the most important message of the study. I-PSS and FV charts should be considered as complementary tools to be used together for an accurate assessment of patients with LUTS. However, although a careful assessment of a patient’s symptoms by questionnaire is considered the first step in the evaluation of men with LUTS suggestive of benign prostatic hyperplasia, the routine use of FV charts in clinical practice as well as in clinical trials has to be established. A better understanding of the relationship between symptom frequency, severity, and bother continues to be challenging and should be considered a key point for the assessment and treatment of LUTS. The authors should be congratulated for their analysis of this important issue.
References [1] Chatelain C, Denis L, Foo JKT, et al. Evaluation and treatment of lower urinary tract symptoms (LUTS) in older men. In: Chatelain C, Denis L, Foo KT, et al., editors. Benign prostatic hyperplasia. Plymouth, United Kingdom: Health Publication; 2001. p. 519–34. [2] Ku JH, Hong SK, Kim HH, et al. Is questionnaire enough to assess number of nocturia episodes? Prospective comparative study between data from questionnaire and frequency–volume charts. Urology 2004;64:965–9. [3] van Venrooij GE, Eckhardt MD, Boon TA. Data from frequency–volume charts versus maximum free flow rate, residual volume, and voiding cystometric estimated urethral obstruction grade and detrusor contractility grade in men with lower urinary tract symptoms suggestive of benign prostatic hyperplasia. Neurourol Urodyn 2002;21:450–6. [4] Yap T, Cromwell DA, Brown C, et al. The relationship between objective frequency–volume chart data and the I-PSS in men with lower urinary tract symptoms. Eur Urol 2007;52:811–8.
DOI: 10.1016/j.eururo.2007.01.014 DOI of original article: 10.1016/j.eururo.2007.01.013
available at www.sciencedirect.com journal homepage: www.europeanurology.com
Benign Prostatic Enlargement – Bladder Outlet Obstruction
The Relationship between Objective Frequency–Volume Chart Data and the I-PSS in Men with Lower Urinary Tract Symptoms Tet L. Yap a,b,*, David A. Cromwell a, Christian Brown a, Jan van der Meulen a, Mark Emberton a,b a b
Clinical Effectiveness Unit, Royal College of Surgeons of England, London, UK Institute of Urology and Nephrology, University College Hospital, London, UK
Article info
Abstract
Article history: Accepted January 4, 2007 Published online ahead of print on January 12, 2007
Objective: The International Prostate Symptom Score (I-PSS) is considered a benchmark in defining the severity of lower urinary tract symptoms (LUTS). However, its relationship with variables of the frequency–volume (FV) chart, an objective measure of LUTS, is not well understood. We analyzed this relationship between the I-PSS and FV chart variables in men with LUTS. Methods: Patients referred for the investigation of uncomplicated LUTS completed an I-PSS and a 3-d FV chart. Correlation coefficients were used to describe the relationship between total I-PSS and its components with the FV variables. The distributions of FV variable scores within the categories of I-PSS severity were also examined. The effect of patient’s quality of life score on I-PSS and FV variables was analyzed using regression. Results: A total of 104 patients completed both the I-PSS and FV chart. The association between FV variables and I-PSS scores was generally weak, with 24-h frequency being the strongest (r = 0.43) and with considerable overlap within each severity category. There was weak association between the I-PSS for nocturia and urgency and their FV counterparts. The number of selfreported nocturnal voids tended to be overestimated. These discrepancies were partly explained by the patient’s quality of life rating, which reduced the level of variation between I-PSS score and FV variables and accounted for some of the overestimation. Conclusion: Weak associations were found between the self-rated I-PSS scores and FV measures of voiding, suggesting that the accurate assessment of LUTS requires self-reported measures of symptoms and objective measures of voiding behaviour. Significant differences between the two types of data might be explained by perceived quality of life.
Keywords: AUA Quality of Life Score Benign prostatic hyperplasia Frequency–volume charts International Prostate Symptom Score Lower urinary tract symptoms
# 2007 European Association of Urology. Published by Elsevier B.V. All rights reserved. * Corresponding author. Clinical Effectiveness Unit, Royal College of Surgeons of England, 35–43 Lincoln’s Inn Fields, London WC2A 3PE, UK. Tel. +44 077 806 11631; Fax: +44 020 7869 6644. E-mail address: [email protected] (T.L. Yap). 0302-2838/$ – see back matter # 2007 European Association of Urology. Published by Elsevier B.V. All rights reserved.
doi:10.1016/j.eururo.2007.01.013
812
1.
european urology 52 (2007) 811–818
Introduction
Lower urinary tract symptoms (LUTS) affect more than 40% of men over 60 yr of age in the United States and Europe [1,2]. The majority of these symptoms are attributed to benign prostatic hyperplasia (BPH), but they can also arise from other causes [3]. It is now common to distinguish between three types of LUTS: storage symptoms, such as frequency and urgency; voiding symptoms, such as slow stream; and postmicturition symptoms, such as incomplete emptying [4]. The International Prostate Symptom Score (I-PSS), a questionnaire that is widely used to assess LUTS in patients, covers all three types, measuring incomplete emptying, frequency, intermittency, urgency, weak stream, hesitancy, and nocturia. A final question (the AUA Quality of Life Score, AUAQoL) asks about how the symptoms affect the patient’s quality of life. A patient gives each symptom a rating from 0 to 5 based on its severity over the past month. These are summed to produce an overall severity score, which is typically used to categorize patients as having mild (score 0–7), moderate (8–19), or severe (20–35) symptoms [5]. However, there has been increasing debate on the precision and accuracy of the I-PSS. First, patient recall may introduce errors. Second, voiding behaviour varies naturally over time, and about one in five patients might be expected to have their total I-PSS score fluctuate by an average of five points, when measured within 1 mo [6]. This can lead to changes in the categorization of LUTS severity in more than one-quarter of men with BPH [7]. Third, symptom questionnaires have been shown to significantly overestimate the severity of symptoms like nocturia and frequency [8,9]. The implications of these findings are that treatment decisions and conclusions about treatment effectiveness based on the I-PSS are more arbitrary than they might appear [7]. A different measure of voiding behaviour is the frequency–volume (FV) chart. This noninvasive instrument is likely to be less susceptible to recall bias and allows the assessment of LUTS outside the artificial setting of a clinic. These charts have been used since the 1970s to investigate LUTS [10], as well as to monitor the outcomes of therapeutic interventions such as bladder training [11]. The precision of chart-recorded data in describing storage symptoms like nocturia, frequency, and urgency have prompted many to recommend its wider use in clinical practice [5,8,11,12]. Despite their widespread use, the relationship between the different FV chart variables and the
I-PSS is not well understood. For instance, it has been noted that frequency measured using an FV chart is consistency lower than patients’ responses to a symptom questionnaire, and it has been suggested that patients who are more disturbed by their symptoms may tend to overestimate their condition when answering a question [11]. The distribution of values for FV variables within I-PSS severity groups (mild, moderate, severe) have yet to be examined. Also, the relationship between individual I-PSS components and related FV variables remains unclear. In this study, we investigated the relationship between the I-PSS and FV chart variables in men with LUTS. We analyzed the association between FV variables and the total I-PSS score, as well as the association between selected FV variables and individual I-PSS questions. Finally, we examined whether the quality of life question of the I-PSS can explain any differences in the severity of symptoms measured using the I-PSS and the FV chart. 2.
Methods
Between February 2003 and June 2004, 140 men with uncomplicated LUTS were recruited from the outpatient department of two urological centres in London. All patients over 40 yr of age who were referred for the first time by their general practitioner to one of the two participating urological outpatient departments with LUTS were eligible for inclusion. Men were excluded: (1) if they had received any form of medical therapy (a-blocker, 5-a reductase inhibitor, or anticholinergic) or had prostatic surgery or pelvic radiotherapy within the prior 3 mo; (2) if they had severe symptoms necessitating immediate medical or surgical treatment; (3) if they had complications potentially related to their symptoms (PSA over 4 ng ml 1, residual volumes over 200 ml, creatinine above 130 mmol l 1, bladder stones, haematuria, urinary retention, recurrent urinary tract infections); (4) if they were unable to speak or understand English; or (5) if they had uncontrolled diabetes, dementia, or end-stage cardiac or respiratory failure. Patients completed an I-PSS questionnaire during their outpatient visit after they had been enrolled in the study and were given a 3-d FV chart to complete within 14 d. On the chart, patients recorded their waking time and time to bed for 3 d, along with volume of liquids drunk to the closest hour and volume of urine voided to the closest hour (measured using a calibrated jug). The parameters extracted from the FV charts were symptom-related measures, such as 24-h frequency, nocturnal frequency, daytime frequency, and number of urgent episodes, and volume-related measures, such as 24-h total intake, 24-h total voided volume, mean voided volume, maximum voided volume, and nocturnal volume. All were calculated as defined by the International Continence Society [4], with frequency and mean volume variables being expressed per 24 h.
813
0.17 0.05 0.04 0.22 0.11 0.17 0.21 0.38* 0.12 0.36* 0.23 0.28* 0.17 0.27* 0.17 0.20 0.24 0.15 0.36* 0.28* 0.28* 0.18 0.18 0.11 0.13 0.11 0.04 0.20 0.18 0.13 0.17 0.20 0.11 0.21 0.08 0.10
0.05 0.05 0.04 0.01 0.08 0.24 0.07 0.03 0.11
0.02 0.02 0.02 0.06 0.08 0.24 0.08 0.06 0.11
0.41* 0.28* 0.22 0.20 0.15 0.12 0.15 0.23 0.04
0.22 0.06 0.04 0.08 0.04 0.09 0.12 0.43* 0.06
Nocturnal volume per 24 h Episodes of urgency per 24 h Episodes of nocturia per 24 h Daytime frequency 24-h frequency Maximum voided volume Mean voided volume Total voided volume per 24 h
Correlation (either g or r) is significant at 0.01 level.
All 140 patients completed the I-PSS questionnaire; 104 men completed at least 1 d of the FV chart, and
*
Results
Total I-PSS Incomplete emptying (Q1) Frequency (Q2) Intermittency (Q3) Urgency (Q4) Weak stream (Q5) Straining (Q6) Nocturia (Q7) AUA-QoL score
3.
Symptom
The general characteristics of the enrolled patients were described using summary statistics (ie, mean and SD). The direct relationship between the total I-PSS score and the FV variables was analyzed using Pearson correlation. The association between the quality of life rating and the FV variables was analyzed using Spearman rank correlation due to the short scale of the AUA-QoL. Analysis of variance was then used to determine whether there were systematic differences between distributions of the FV variables and the categories of I-PSS severity. The total I-PSS score was categorized in the standard way, using 0–7 to denote mild symptoms, 8–19 to denote moderate symptoms, and 20–35 to denote severe symptoms [13]. Box plots were used to show the distribution of values for those variables whose mean differed across the I-PSS categories. Because the total I-PSS score was the sum of the individual question scores, multiple linear regression was used to assess whether the total I-PSS score was related to a combination of FV variables. Stepwise regression was used to find the most parsimonious model. The significance level for removing variables was set at p = 0.2 [14]. The relationship between individual I-PSS questions and FV variables was analyzed using Spearman rank correlation and the gamma statistic (g). The gamma statistic is preferable to rank correlation if the data contain many tied observations because both variables are measured on short scales [14]. Like other measures of association, the gamma statistic ranges from 1 to 1 and is 0 when there is no association. The analysis focussed on those FV variables expected to be strongly related to particular I-PSS questions. The 24-h frequency was compared to the I-PSS questions on incomplete emptying and frequency. The number of urgent voids was compared to the I-PSS questions on urgency and frequency. Finally, the FV variables for the nocturnal frequency and nocturnal volume were compared to the I-PSS question on nocturia. Both instruments define nocturia as the number of nocturnal voids. This enabled us to examine whether selfreported values were overestimated. Multiple regression was used to assess the degree of underreporting and whether this might be explained by perceived quality of life as measured using the AUA-QoL. All other I-PSS questions ask patients to rate symptom burden as: 0 for ‘‘not at all,’’ 1 for ‘‘less than one in five times,’’ 2 for ‘‘less than half the time,’’ 3 for ‘‘about half the time,’’ 4 for ‘‘more than half the time,’’ and 5 for ‘‘almost always.’’ This scale was not equivalent to the scales used by the FV variables, and this difference could weaken the association between similar questions. In the case of urgency, this possibility could be examined by transforming the variable into an equivalent scale. First, the percentage of the total voids that were urgent was calculated. The transformed variable was then defined based on increasing 20% point intervals (0% was assigned a score of 0, 1–20% was assigned a score of 1, and so on).
Total intake volume per 24 h
Statistical analysis
Table 1 – Relationship between individual frequency–volume variables and I-PSS
2.1.
0.04 0.09 0.18 0.03 0.19 0.20 0.01 0.15 0.13
european urology 52 (2007) 811–818
814
european urology 52 (2007) 811–818
Table 2 – Mean and standard deviation of frequency–volume chart variables by I-PSS severity category (mild, 0–7; moderate, 8–19; severe, 20–35) Frequency–volume chart variable 24-h frequency Daytime frequency Nocturnal frequency Number of urgent episodes Total voided volume per 24 h Mean voided volume Maximum voided volume Nocturnal volume per 24 h Total intake volume per 24 h
Mild I-PSS (n = 6) mean (SD)
Moderate I-PSS (n = 76) mean (SD)
Severe I-PSS (n = 22) mean (SD)
9.1 (3.5) 7.6 (3.4) 1.5 (1.1) 1.2 (2.3) 1573 (736) 172 (32) 408 (66) 510 (195) 1990 (1110)
9.1 (2.2) 7.4 (2.0) 1.7 (0.9) 2.1 (2.6) 1707 (671) 188 (62) 372 (150) 584 (277) 1891 (750)
11.0 (3.3) 8.7 (2.8) 2.2 (1.3) 3.8 (3.7) 1848 (723) 172 (62) 376 (177) 683 (333) 1918 (602)
96 patients completed all 3 d. Among the men completing both instruments, the mean age was 63 yr (SD = 11, range: 40–86) and the mean symptom duration was 4.4 yr (SD = 6.2). Most men were Caucasian (81 of 104, 78%) and had some academic qualification (75 of 104, 72%). The mean scores on the I-PSS and AUA-QoL were 16.2 (SD = 6.1) and 3.6 (SD = 1.1), respectively. There was no important difference in age, ethnicity, level of education, mean symptom duration, I-PSS symptom, or AUA-QoL scores between the 104 men who supplied frequency–volume data and the 36 men who did not. The associations between total I-PSS score and the FV variables were highest for the variables derived from the number of voids (Table 1). The 24-h frequency (r = 0.41) and number of urgent episodes (r = 0.36) showed the largest (although only moderate) correlations. Daytime frequency and total I-PSS showed a similar correlation (r = 0.36), but the daytime values were strongly correlated with 24-h frequency (r = 0.95). There was no association between total I-PSS and the variables based on volume voided. There was also no association between AUA-QoL scores and any of the FV variables. The mean values of the different FV variables did not exhibit a strong trend across the categories of I-PSS severity (Table 2). Indeed, consistent increases were observed for only four variables, and the differences between severity groups were only statistically significant for the mean number of voids per 24 h and episodes of urgency. For both these variables, however, the distribution of FV values within each I-PSS severity category overlapped substantially (Figs. 1, 2), demonstrating that the I-PSS severity categories do not discriminate effectively between patients with different voiding behaviours. In the regression model developed to assess whether total I-PSS could be predicted from a combination of all FV chart variables, only the coefficients for 24-h frequency and urgency were
ANOVA F-test p value 0.01 0.06 0.06 0.04 0.59 0.50 0.85 0.26 0.95
statistically significant at a 5% level, being 0.76 (95% CI: 0.33–1.12) and 0.55 (95% CI: 0.16–0.95), respectively. The model explained only 22% of the variation in total I-PSS.
Fig. 1 – Distribution of 24-h frequency from frequency– volume chart data by I-PSS severity category (mild, 0–7; moderate, 8–19; severe, 20–35).
Fig. 2 – Distribution of urgent voids from frequency–volume chart data by I-PSS severity category (mild, 0–7; moderate, 8–19; severe, 20–35).
815
european urology 52 (2007) 811–818
Table 3 – Mean and standard deviation of frequency–volume chart-recorded variables aggregated by scores on related I-PSS questions FV chart variable
I-PSS question
I-PSS question score 0
24-h frequency
1
2
Correlation
3
4
5
Q1 (incomplete emptying) Q2 (frequency)
8.9 (2.0)
8.6 (2.2)
9.3 (2.4)
10.4 (3.3)
10.3 (2.6)
12.9 (2.3)
r = 0.25
9.3 (0.5)
7.6 (2.2)
8.9 (2.0)
9.5 (2.9)
9.8 (1.9)
10.9 (3.6)
r = 0.33*
Number of urgent episodes
Q2 (frequency) Q4 (urgency) Q4 (urgency) transformed
2.8 (4.0) 1.9 (2.9) 1.3 (1.7)
1.3 (1.5) 1.5 (2.0) 1.1 (1.2)
1.8 (3.2) 2.0 (2.8) 1.3 (1.4)
1.9 (2.4) 2.1 (2.5) 1.4 (1.5)
2.8 (2.5) 1.9 (2.3) 1.4 (1.4)
4.1 (4.2) 5.5 (3.8) 2.9 (1.4)
r = 0.29* r = 0.29* g = 0.27
Nocturnal frequency Nocturnal volume (ml)
Q7 (nocturia) Q7 (nocturia)
0.6 (0.1) 513 (53)
1.2 (0.7) 465 (238)
1.6 (0.8) 538 (165)
2.0 (0.9) 645 (326)
2.3 (0.8) 718 (269)
3.2 (1.9) 917 (452)
r = 0.44* r = 0.32*
*
Correlations are significant at 0.01 level.
Table 3 summarizes the association between the individual I-PSS questions and the FV variables. Most FV variables did not show any significant relationship with their I-PSS counterparts. The association between I-PSS question 4 (urgency) and the equivalent FV variable was weak (r = 0.29). This did not appear to be due to the different scales, because the association was also weak between the I-PSS question and the transformed urgent FV variable (g = 0.27). Indeed, the effect of the transformation was small and the normal and transformed urgent FV variables were highly correlated (Spearman’s r = 0.97). There was a positive association between nocturnal frequency and the I-PSS nocturia question, although the agreement was not strong (g = 0.43). The self-reported nocturnal voids tended to be higher than those actually recorded on the FV chart (Fig. 3), with the discrepancy increasing as the number of self-reported voids increased. Patients who reported no nocturnal voiding on the I-PSS were
Fig. 3 – Distribution of nocturnal frequency from frequency–volume chart data across the scores on I-PSS question 7 (nocturia).
the exception, with the majority recording some nocturnal voids on their FV chart. Regression was used to examine whether the difference between the I-PSS and FV-chart nocturnal frequency was related to perceived quality of life due to a patient’s urinary symptoms. The predicted I-PSS was estimated to increase by only 0.5 voids for every unit increase in FV nocturnal frequency (95% CI: 0.3–0.7). The predicted I-PSS was estimated to increase by 0.3 for every unit increase in the AUA-QoL variable, suggesting differences in quality of life ratings explains some of the overestimation. Nonetheless, the regression model could only explain 31% of the variation between patients in I-PSS nocturnal voids.
4.
Discussion
This study showed that the total I-PSS correlated moderately with FV chart symptom measures of frequency and urgency but not with the FV volume variables. When patients were categorized according to their I-PSS severity, the mean severity of frequency and urgency FV variables increased across the mild, moderate, and severe groups. However, there was a large overlap in the distributions of FV values between these groups, and knowing a patient’s individual value would not allow a clinician to accurately infer what their severity category might be. The poor association between individual variables and the total I-PSS did not seem to be due to the aggregate construction of the total I-PSS. When we assessed how well a combination of all FV variables could predict the total I-PSS, only 24-h frequency and urgency were found to explain some of the observed variation in total I-PSS. The correlation between individual FV variables and related I-PSS questions was also found to range
816
european urology 52 (2007) 811–818
from weak to moderate. The different scales used by the I-PSS questions and the FV variables may be a source of variation that limited the correlation; in the case of the two measures of urgency, however, transforming the FV variable to an equivalent scale did not improve the correspondence between the measurements. The higher scores on I-PSS question for nocturia significantly overestimated chart recordings, which was consistent with other studies of patients referred to urological clinics [8,9,15]. It has been hypothesized that patients who are more disturbed by their symptoms may tend to overestimate their condition [11]. Our results seem to support this. The AUA-QoL score gives an indication of how urinary symptoms affect a patient’s quality of life and was found to partially explain some of the overestimation. The I-PSS correlates poorly with certain objective measures of urinary symptoms, such as mean and maximum flow rates, prostate size, and postvoid residual volume [16]. Few other studies have examined the relationship between FV chart data and the I-PSS in patients with LUTS. Ku et al. reported poor agreement between total I-PSS and FV chart variables in 164 patients referred for LUTS evaluation, noting that many overestimated their nocturia on I-PSS [17]. Another study of 160 men with LUTS found some correlation between small voided volumes (Kendall and Gibbons coefficient, r = 0.17), nocturia (r = 0.21), and daytime frequency (r = 0.31) with total I-PSS [15]. That study used only 24-h FV charts, however, which have been shown to produce less reliable values due to the amount of day-to-day variation in voiding behaviour [5,12]. The lack of correlation between the two instruments seen with our data may be due to a number of study limitations. First, only 74% of patients completed at least 1 d of the FV charts. Nevertheless, this compliance rate is similar to rates reported by other studies using FV charts [9,17]. Moreover, there were no differences in the characteristics of patients, their total I-PSS, or the AUA-QoL score between responders and nonresponders ( p > 0.5 for all). Second, the weak correlation could be due to poor chart recording. To reduce this risk, all subjects were taught to fill the charts by a specialist nurse. Analysis was performed using data from patients who completed charts for 1, 2, and 3 d to ensure the analyzed sample was as representative as possible. In fact, it appeared that this group of patients did not differ significantly from patients who completed charts for all 3 d, and repeating the analysis using only data from 3-d charts produced equivalent results. Third, the group of men enrolled in the study may be more homogenous than patient
populations seen normally in clinics. This would cause the correlation coefficients to be underestimated. However, the range of scores observed in our sample was similar to those reported elsewhere [18,19] and does not seem to be a substantial source of bias. Fourth, the patients completed the FV chart after being seen in the clinic, whereas patients’ I-PSS answers refer to their symptoms over the past month. Consequently, adjacent periods in time are being compared and it is possible that the poor agreement was due to changes in patients’ underlying conditions. This seems unlikely, however, as LUTS is a chronic condition, and the patient cohort did not report any adverse consequences such as infections or acute retention during their clinical assessment. Finally, although no actual medical or surgical intervention was performed, the enrolled patients saw a medical professional in the assessment clinic. This may have influenced their perception of their symptoms and consequently the prospectively collected FV data. However, there is no evidence to suggest that a single visit would result in such large changes. The differences in the information provided by the two measures of LUTS suggest that the two are complementary assessment tools. The Fifth International Consultation on BPH recommended the use of both the I-PSS (with the AUA-QoL question) and FV charts (especially in patients with nocturia) in initial LUTS evaluation [20]. The lack of association between the I-PSS and frequency measures suggests that the FV chart is better suited to providing information regarding the aetiology of symptoms. For instance, mean voided volume can be reduced by a decrease in functional bladder capacity (eg, due to overactive bladder disorders). Also, FV volume data such as nocturnal volume are important in distinguishing causes of nocturia, including nocturnal polyuria [21]. Another reason to use both tools is the apparent difference in self-reported and actual voiding behaviour. The I-PSS for nocturia overestimated chart recordings, and using both tools allows for objective and perceived levels of symptom severity to be differentiated, both when a patient is assessed and when monitoring the effect of treatment. Our results suggest that the severity of nocturia reported by patients is influenced by the way patients perceive their quality of life due to their urinary symptoms, and this may also have an influence on their health-care seeking behaviour. A trial using a population-based sample showed an underestimation of nocturia compared to FV measurements [8]. The way in which men’s experiences and perceptions may influence their self-reported symptoms is
european urology 52 (2007) 811–818
an area that requires further investigation. The AUA-QoL is a single question embodying a specific interpretation of quality of life, namely, the perceived burden of urinary symptoms. However, quality of life is a complex concept that covers bother, physical and mental health, functional impairment, and social integration. Bother reflects the overall distress of having LUTS and is strongly related to social anxiety. Furthermore, LUTS affects patient’s self-perception, leading to restrictions in social activities [22]. Validated instruments exist to measure these dimensions of quality of life, including condition-specific measures such as the BPH Impact Index, the DAN-PSS, and the ICSQoL [23,24]. We suggest that the apparent overestimation of selfreported symptoms could be better explained if data from these measures could be incorporated into an analysis of the relationship between the I-PSS and FV chart variables.
Conflicts of interest We acknowledge (1) that all authors have made a substantial contribution to the information or material submitted for publication; (2) that all have read and approved the final manuscript; (3) that we have no direct or indirect commercial financial incentive associated with publishing the article.
References [1] Speakman MJ, Kirby RS, Joyce A, Abrams P, Pocock R. Guideline for the primary care management of male lower urinary tract symptoms. BJU Int 2004;93:985–90. [2] Platz EA, Smit E, Curhan GC, Nyberg LM, Giovannucci E. Prevalence of and racial/ethnic variation in lower urinary tract symptoms and noncancer prostate surgery in U.S. men. Urology 2002;59:877–83. [3] Jones D. Benign prostatic hypertrophy and lower urinary tract dysfunction. In: Weiss RM, George NRG, O’Reilly PH, editors. Comprehensive urology. London: Mosby International; 2001. p. 452. [4] Abrams P, Cardozo L, Fall M, et al. The standardisation of terminology of lower urinary tract function: report from the Standardisation Sub-committee of the International Continence Society. Neurourol Urodyn 2002;21:167–78. [5] Palnaes HC, Klarskov P. The accuracy of the frequency– volume chart: comparison of self-reported and measured volumes. Br J Urol 1998;81:709–11. [6] Barry MJ, Girman CJ, O’Leary MP, et al., Benign Prostatic Hyperplasia Treatment Outcomes Study Group. Using repeated measures of symptom score, uroflowmetry and prostate specific antigen in the clinical management of prostate disease. J Urol 1995;153:99–103.
817
[7] O’Connor RC, Bales GT, Avila D, Gerber GS. Variability of the International Prostate Symptom Score in men with lower urinary tract symptoms. Scand J Urol Nephrol 2003;37:35–7. [8] McCormack M, Infante-Rivard C, Schick E. Agreement between clinical methods of measurement of urinary frequency and functional bladder capacity. Br J Urol 1992;69:17–21. [9] Blanker MH, Bohnen AM, Groeneveld FP, Bernsen RM, Prins A, Ruud Bosch JL. Normal voiding patterns and determinants of increased diurnal and nocturnal voiding frequency in elderly men. J Urol 2000;164:1201–5. [10] Torrens MJ. The effect of selective sacral nerve blocks on vesical and urethral function. J Urol 1974;112:204–5. [11] Abrams P, Klevmark B. Frequency volume charts: an indispensable part of lower urinary tract assessment. Scand J Urol Nephrol Suppl 1996;179:47–53. [12] Bryan NP, Chapple CR. Frequency volume charts in the assessment and evaluation of treatment: How should we use them? Eur Urol 2004;46:636–40. [13] Barry MJ, Fowler Jr FJ, O’Leary MP, et al. The Measurement Committee of the American Urological Association. The American Urological Association symptom index for benign prostatic hyperplasia. J Urol 1992;148: 1549–57. [14] Siegel S, Castellan Jr NJ. Measures of association and their tests of significance. In: Siegel S, Castellan Jr NJ, editors. Nonparametric statistics for the behavioural sciences. New York: McGraw Hill; 1988. p. 291–3. [15] Van Venrooij G-EPM, Eckhardt MD, Gisolf K-WH, Boon TA. Data from frequency–volume charts versus symptom scores and quality of life score in men with lower urinary tract symptoms due to benign prostatic hyperplasia. Eur Urol 2001;39:42–7. [16] Wadie BS, Ibrahim EH, de la Rosette JJ, Gomha MA, Ghoneim MA. The relationship of the International Prostate Symptom Score and objective parameters for diagnosing bladder outlet obstruction. Part I: when statistics fail. J Urol 2001;165:32–4. [17] Ku JH, Hong SK, Kim HH, Paick JS, Lee SE, Oh SJ. Is questionnaire enough to assess number of nocturic episodes? Prospective comparative study between data from questionnaire and frequency–volume charts. Urology 2004;64:966–9. [18] Bertaccini A, Vassallo F, Martino F, et al. Symptoms, bothersomeness and quality of life in patients with LUTS suggestive of BPH. Eur Urol 2001;40(Suppl 1):13–8. [19] Ezz ED, Kiemeney LA, de Wildt MJ, Debruyne FM, de la Rosette JJ. Correlation between uroflowmetry, prostate volume, postvoid residue, and lower urinary tract symptoms as measured by the International Prostate Symptom Score. Urology 1996;48:393–7. [20] Roehrborn CG. Focus on lower urinary tract symptoms: nomenclature, diagnosis, and treatment options—highlights from the 5th International Consultation on Benign Prostatic Hyperplasia, June 25–27, 2000, Paris, France. Rev Urol 2001;3:139–45. [21] Weiss JP. Nocturia: ‘‘do the math’’. J Urol 2006;175: S16–8.
818
european urology 52 (2007) 811–818
[22] Glover L, Gannon K, McLoughlin J, Emberton M. Men’s experiences of having lower urinary tract symptoms: factors relating to bother. BJU Int 2004;94:563–7. [23] Donovan JL, Kay HE, Peters TJ, et al. Using the ICSQoL to measure the impact of lower urinary tract symptoms on quality of life: evidence from the International
Editorial Comment on: The Relationship between Objective Frequency–Volume Chart Data and the I-PSS in Men with Lower Urinary Tract Symptoms Cosimo De Nunzio, Department of Urology, Ospedale Sant’Andrea, Roma, Italy [email protected] According to the 5th International Consultation on BPH [1], recommended tests to be performed in all patients with lower urinary tract symptoms (LUTS) and benign prostatic enlargement (BPE) included medical history, quantification of symptoms and bother, physical examination, urinalysis, prostate-specific antigen, and frequency–volume (FV) chart. FV charts are simple to complete and provide useful clinical information as an estimation of minimum voided volume, mean voided volume, maximum voided volume, number of voids in 24 h, mean diuria, mean voided volume in daytime, mean nocturia, mean voided volume at night; and mean fluid intake in daytime and at night. In patients with nocturia, FV charts are considered the most valid instrument to distinguish between nocturia, nocturnal polyuria, or excessive fluid intake [2]. Data from FV charts has been used to compare a patient’s voiding habits at home and the urodynamic variables assessed by uroflowmetry and filling cystometry [3]. The authors of this report [4] evaluated the relationship between FV charts data and the severity of symptoms and bother assessed by the International Prostatic Symptom Score (I-PSS) in men with LUTS. Overall a weak association was detected between total I-PSS and FV variables with the 24-h frequency (r = 0.41) and the number of urgent episodes (r = 0.36) showing the largest but moderate correlations. No associations between FV variables and quality-of-life score were observed. Most FV variables did not show any significant relationship with their I-PSS counterparts and a positive but not strong association was revealed between nocturnal frequency and the I-PSS nocturia question with the highest scores on the I-PSS question for nocturia, overestimating chart
Continence Society–Benign Prostatic Hyperplasia Study. Br J Urol 1997;80:712–21. [24] Haltbakk J, Hanestad BR, Hunskaar S. Use and misuse of the concept of quality of life in evaluating surgical treatments for lower urinary tract symptoms. BJU Int 2003; 91:380–8.
recordings. The lack of correlation between these two measures of LUTS is the most important message of the study. I-PSS and FV charts should be considered as complementary tools to be used together for an accurate assessment of patients with LUTS. However, although a careful assessment of a patient’s symptoms by questionnaire is considered the first step in the evaluation of men with LUTS suggestive of benign prostatic hyperplasia, the routine use of FV charts in clinical practice as well as in clinical trials has to be established. A better understanding of the relationship between symptom frequency, severity, and bother continues to be challenging and should be considered a key point for the assessment and treatment of LUTS. The authors should be congratulated for their analysis of this important issue.
References [1] Chatelain C, Denis L, Foo JKT, et al. Evaluation and treatment of lower urinary tract symptoms (LUTS) in older men. In: Chatelain C, Denis L, Foo KT, et al., editors. Benign prostatic hyperplasia. Plymouth, United Kingdom: Health Publication; 2001. p. 519–34. [2] Ku JH, Hong SK, Kim HH, et al. Is questionnaire enough to assess number of nocturia episodes? Prospective comparative study between data from questionnaire and frequency–volume charts. Urology 2004;64:965–9. [3] van Venrooij GE, Eckhardt MD, Boon TA. Data from frequency–volume charts versus maximum free flow rate, residual volume, and voiding cystometric estimated urethral obstruction grade and detrusor contractility grade in men with lower urinary tract symptoms suggestive of benign prostatic hyperplasia. Neurourol Urodyn 2002;21:450–6. [4] Yap T, Cromwell DA, Brown C, et al. The relationship between objective frequency–volume chart data and the I-PSS in men with lower urinary tract symptoms. Eur Urol 2007;52:811–8.
DOI: 10.1016/j.eururo.2007.01.014 DOI of original article: 10.1016/j.eururo.2007.01.013