British Journal of Urology (1998), 81, 36–41

Comparison of the Danish prostatic symptom score with theInternational Prostatic Symptom Score, the Madsen-Iversenand Boyarsky symptom indexesB.J. HANSEN 1 , S. MORTENSE N2 , H.J .A. MENSINK 3 , H. FLYGER 4 , M. RIEHMANN 5 , N. HE NDOLIN 6 ,J . NORDLING 5 , T. HALD 5 and the ALFECH Study Gro up*1Department of Urology, Bispebjerg Hospital, Copenhagen, 2Department of Surgery D, Section of Urology, Glostrup Hospital,Denmark, 3Department of Urology, Academic Hospital, Groningen, The Netherlands, 4Department of Surgery A, Section of Urology,Hillerød Hospital, Denmark, 5Department of Urology, Herlev Hospital, Herlev, Denmark, and 6Department of Urology, Eira Hospital,Helsinki, Finland

Objective To compare the Danish Prostatic Symptom were correlated on a pairwise basis. The DAN-PSS andIPSS indexes have significant construct validity inScore (DAN-PSS) with the International Prostatic

Symptom Score (IPSS), Madsen-Iversen and Boyarsky terms of correlation with the Madsen-Iversen system(Spearman’s correlation coeBcient, r

s=0.51 and r

s=symptom indexes in a clinical setting, and to evaluate

the potential significance of any diCerences in 0.45, respectively). The DAN-PSS and the IPSS indexeswere correlated (r

s=0.61). The DAN-PSS was moreinformation obtained from these questionnaires.

Patients and methods The study comprised two sensitive than the IPSS to changes after pharmacologi-cal treatment, with scores decreasing 70% and 29%substudies: in the first, 205 patients with lower urinary

tract symptoms (LUTS) suggestive of bladder outlet (P<0.05), respectively, after treatment with an alpha-blocker for 4 months, and 50% and 29% (P<0.05),obstruction (BOO), a Madsen-Iversen score >6 and a

maximum flow rate of <10 mL/s were randomized to respectively, after 4 months on placebo treatment.Finally, the responsiveness of the Boyarsky and DAN-receive either placebo or alfuzosin in a double-blind

study of 16 weeks. The symptoms were assessed PSS indexes to TUMT showed that the DAN-PSSsystem was significantly more responsive than theusing the Madsen-Iversen, DAN-PSS and the IPSS

questionnaires. In the second, 138 patients with LUTS Boyarsky index, with scores decreasing 57% and 15%(P<0.05), respectively, after one year.suggestive of BOO were selected for treatment with

transurethral microwave thermotherapy (TUMT, 52°C Conclusions The DAN-PSS index is more sensitive thanthe IPSS, Madsen-Iversen and Boyarsky symptomfor 60 min, microwave energy 200 kJ) and their symp-

toms assessed using the Boyarsky and the DAN-PSS indexes, incorporates important outcome events,includes a patient-weighting of each symptom, therebyquestionnaires. Patients were then followed for one

year. Rank correlation coeBcients and regression lines reflecting better the patients’ global assessment ofoutcome.were calculated using Spearman’s non-parametric test.

The relative changes, i.e. responsiveness, calculated Keywords Lower urinary tract symptoms, LUTS, bladderoutlet obstruction, BOO, Danish Prostatic Symptomfor the DAN-PSS, IPSS and Boyarsky indexes were

compared pairwise using the Wilcoxon-Pratt test. Score, DAN-PSS, IPSS, Madsen-Iversen, Boyarsky,correlation, comparison, sensitivityResults The DAN-PSS, IPSS and Madsen-Iversen indexes

toms, and thereby improvement in his quality of life,Introductionrather than improvements in flow rate, detrusor pressureor other urodynamic factors [2]. Given this centralThe indications for the treatment of LUTS suggestive of

BOO are relative and a careful analysis of the symptoma- role of LUTS, it is important to assess accurately theoccurrence, progression and/or resolution of LUTStology appears crucial in counselling patients [1].

Furthermore, the single most important outcome of suggestive of BOO. To maintain conformity among clini-cal investigators several symptom scoring systems havetherapy for the patient is the relief of bothersome symp-

beendeveloped, including the Boyarsky [3] and Madsen-Accepted for publication 24 September 1997

*Members of the ALFECH study group are listed in Appendix 1. Iversen [4] (Table 1). Both systems rate the symptoms

36 © 1998 British Journal of Urology

COMPARISON OF SYMPTOM SCORES 37

Table 1 Symptoms included in clinically used symptom score systems

Symptom score Boyarsky (1977) Madsen-Iversen (1983) DAN-PSS (1991) AUA-7 (1992) IPSS (1991, 1993, 1995)

Daytime frequency 0–3 0–3 0–3 0–5 0–5Nocturia 0–3 0–3 0–3 0–5 0–5Urge 0–3 0–3 0–3 0–5 0–5Urge incontinence 0–3Hesitancy 0–3 0 or 3 0–3Poor flow 0–3 0–4 0–3 0–5 0–5Incomplete emptying 0–3 0–4 0–3 0–5 0–5Straining 0 or 2 0–3 0–5 0–5Intermittency 0–3 0 or 3 0–5 0–5Dysuria 0–3 0–3Post-micturition dribbling 0–3 0 or 2 0–3Stress incontinence 0–3Overflow/seeping incontinence 0–3Bother (QOL) questions 0–3 0–3 or 0–6 0–6

for each Q# two Q# one Q#

referred to as ‘obstructive’ higher than those referred to naire provides information related to aetiologyor pathophysiology of the urinary symptoms. Theas ‘irritative’. These questionnaires are not designed to

be self-administered, which may introduce a potential AUA system includes an additional question on theglobal impact of LUTS on quality of life and wasbias from the interviewer depending on diCerent inter-

viewing techniques. Additionally, neither of these symp- adopted as the IPSS. The two subsequent conferences,in 1993 and 1995, made no revisions, althoughtom scores, although sensitive to clinical changes over

time, identify the bothersomeness of symptoms and the AUA system, and therefore the IPSS, has beencriticized.possible disease impact on quality of life, and neither has

been studied to determine their reliability and validity. The DAN-PSS seems more extensive than the IPSS(Table 1); one of the major diCerences between them isAs part of a multifactorial treatment strategy for BPH

the International Consensus Conference in Paris, 1991 that the former evaluates the symptoms both quantitat-ively and qualitatively by determining both a symptom[2] recommended that future symptom-score systems

should include a patient weighting of the symptoms in and a bother score, whereas the IPSS does not includea section on the bother of each urinary symptom. Theterms of the impact on the quality of life of the

patient. The AUA symptom-score system and the objective of the present study was to compare the DAN-PSS with the IPSS, Madsen-Iversen and BoyarskyDanish Prostatic Symptom Score (DAN-PSS) system are

the most recently validated [5–7] (Table 2). Both were symptom indexes in a clinical setting, and to evaluatethe potential significance of diCerences on the infor-designed as objective, reproducible tools to measure

symptoms and response to therapy; neither question- mation obtained from these questionnaires.

Table 2 The reliability and validity of theDAN-PSS symptom score system and theAUA symptom score system

Variable DAN-PSS AUA

Internal consistency (Cronbach’s alpha) 0.73 0.85Test/re-test reliability 0.84 0.81Construct validity:Correlation with Madsen-Iversen score (r

s) 0.51 0.85

Correlation with ‘bother’ (rs) 0.71 0.61

Discriminant validity (area under the ROC curve) 0.94 0.87Responsiveness after TURP:

Baseline score, median (quartiles) 20 (14–36)Absolute decrease* (4 months) 20% decrease 100Absolute decrease* (6 months) 19% decrease 95 71

*Absolute decrease from median baseline score at 12–16 weeks after TURP

38 B.J. HANSE N et al.

Iversen system. The regression line and the regressionPatients and methodsequation (Fig. 1b, d) provide a method to approximateequivalent scores between scoring systems; the scatterThe study comprised two substudies. In the first, 205

patients with LUTS suggestive of BOO and no previous of points around the regression line provides an assess-ment of the range of values on another score that areprostatectomy were randomized to receive placebo and

alfuzosin in a double-blind study (the ALFECH study [8]). consistent with a measured value on a known score.Such a comparison is valuable when considering theAfter randomization, the patients received placebo in a

single-blind phase for 4 weeks, followed by alfuzosin thresholds that have been proposed to separate mildfrom moderate and severe symptoms for the DAN-PSS2.5 mg three times daily or placebo for a further 12

weeks. A Madsen-Iversen score of >6 [4] and a maxi- and IPSS indexes. However, the thresholds for the DAN-PSS model have yet to be determined to make such amum flow rate of <10 mL/s were used as inclusion

criteria. The patients completed the DAN-PSS and the comparison. The DAN-PSS and IPSS indexes were weaklycorrelated (Fig. 1b, d).IPSS questionnaires at the time of inclusion (–4 weeks),

as well as at baseline (time 0), 6 and 12 weeks after The responsiveness to pharmacological treatment forthe DAN-PSS and IPSS is shown in Fig. 2. A pairwiseinitiation of the study.

In the second substudy, 138 patients with LUTS comparison of the relative diCerences in the DAN-PSSand IPSS indexes showed that the former had a statisti-suggestive of BOO and no previous prostatic surgery

were treated with the ProstaLund (Lund Instruments AB, cally significantly higher relative response to bothplacebo and alpha-blocker treatment than had the IPSSSweden) computer-controlled transurethral microwave

thermotherapy (TUMT) device at Eira Hospital, Finland, system (P<0.05). Furthermore, the decrease in theDAN-PSS index was statistically greater in the alfuzosinat a treatment temperature of 52°C for 60 min (micro-

wave energy 200 kJ). The patients’ symptoms were than in the placebo group (P<0.05), a diCerence notdetected by the IPSS index. The sensitivity of theassessed by the Boyarsky and DAN-PSS

questionnaires, the patients being followed for one year Boyarsky and DAN-PSS indexes to TUMT (Fig. 3) showedthe latter to be significantly more responsive than the[9].

Rank correlation coeBcients and regression lines were former (P<0.05).calculated using Spearman’s non-parametric test. As theDAN-PSS, the IPSS and the Boyarsky indexes have DiscussiondiCerent ranges, the changes in score are presented asrelative changes, expressed as the percentage of the In evaluating the eCectiveness of diCerent treatments for

BPH and the natural course of the disease, a reliable,pretreatment median values, to allow comparison. Therelative changes, i.e. responsiveness, calculated for the reproducible, internally consistent and clinically relevant

measurement of symptom severity is essential [2]. TheDAN-PSS, IPSS and Boyarsky indexes were comparedpairwise using the Wilcoxon-Pratt test. DAN-PSS and IPSS indexes have acceptable measures of

reliability and validity [5–7] (Table 2) and both symptomquestionnaires are responsive in patients treated pharma-Resultscologically and in those undergoing prostatectomy[11–13]. Lepor and Machi [14] and Chai et al. [15] haveFigure 1 presents selected correlations among the

DAN-PSS, IPSS and Madsen-Iversen indexes on a pair- shown that men and women score equally using theIPSS questionnaire. It has been suggested that LUTS iswise basis. To assess the correlation between the DAN-

PSS and the Madsen-Iversen system, and the IPSS and a manifestation of ageing and that the IPSS is not specificfor BPH. The DAN-PSS has not yet been applied to age-the Madsen-Iversen system, the questions concerning

bother in the DAN-PSS and IPSS systems were excluded matched men and women. To evaluate the discriminativeability of any BPH questionnaire, the symptom score(Fig. 1a, b). The correlation between the IPSS and the

Madsen-Iversen index is shown in Fig. 1(c) and Fig. 1(d) must be administered to men with and without BOO.Yalla et al. [16] have shown, using video-urodynamicthe correlation between the DAN-PSS total score (includ-

ing the bother factor for each question) and IPSS total studies, that the IPSS system could not discriminatebetween men with and without BOO, or with andscore (including the question of the global impact of BPH

on quality of life). without detrusor instability. Based on simultaneouspressure and flow measurements, Poulsen et al. [17]The construct validity of a questionnaire relates to

how well the set of questions correlate with the theoreti- recently found that the DAN-PSS system could notdiCerentiate patients with BOO from unobstructedcal clinical variable [10]. Figure 1(a, c) show that the

DAN-PSS and IPSS indexes have significant construct patients. It is important to emphasize that both theIPSS and DAN-PSS indexes were designed as objective,validity as assessed by correlation with the Madsen-

COMPARISON OF SYMPTOM SCORES 39

16

0

IPSS total symptom score

Mad

sen

-Ive

rsen

sco

re

435

6

8

10

12

14

305 10 15 20 25

c

16

0

DAN-PSS-1 total symptom score

Mad

sen

-Ive

rsen

sco

re

035

2

4

6

8

10

12

14

305 10 15 20 25

a

40

0

DAN-PSS-1 score

IPS

S s

core

060

5

10

15

20

25

30

35

5010 20 30 40

d

35

0

DAN-PSS-1 total symptom score

IPS

S t

ota

l sym

pto

m s

core

030

5

10

15

20

25

30

5 10 15 20 25

b

Fig. 1. Correlation of, a, the DAN-PSS total symptom score with the Madsen-Iversen index (regression line y=0.26x+5.8, rs=0.505,

P<0.001); b, the DAN-PSS with the IPSS (total) (regression line y=0.76×+6.99. rs=0.592, P<0.001); c, the IPSS (total) with the

Madsen-Iversen index (regression line y=0.14×+6.55, rs=0.449, P<0.001; and d, the DAN-PSS with the IPSS (total) (regression line

y=0.38×+11.2, rs=0.608, P<0.001.

100

0

Months

%

012

20

40

60

80

61 3

Fig. 3. The pecentage decline in the DAN-PSS total (red) and theBoyarsky score (green) after ‘open-label’ TUMT. The median(quartiles) baseline absolute values for the DAN-PSS and Boyarskyscores were 14 (8–25) and 10 (8–13), respectively.

100

Baseline

Weeks

%

012

10

20

30

40

50

60

70

80

90

6

Fig. 2. The percentage decline in the DAN-PSS and the IPSS total indexes as evaluative tools, but emphasizes that carescores during placebo-controlled alfuzosin therapy. The median

must be taken to avoid relying solely on the index for(quartiles) baseline absolute values for the IPSS and DAN-PSS werediagnosing the aetiology of LUTS.14 (7–19.5) and 5 (0–11) for the alfuzosin group, and 12 (7–18)

The DAN-PSS index discriminated well between menand 6 (0–14) for the placebo group, respectively. Alfuzosin (IPSS),dark green. Placebo (IPSS), light green. Alfuzosin (DANPSS), light with LUTS suggestive of BOO, who stated that they hadred. Placebo (DANPSS), dark red. general voiding problems, and an age-matched control

group with no voiding problems [7]. The IPSS index alsohas discriminative validity [5], but the discriminativereproducible tools to measure symptoms and longitudinal

changes in symptoms during intervention, and not for ability of this system was determined in a substantiallyyounger control group who completed the questionnaire.diagnosing BOO. Consequently, based on these scoring

systems, it is not possible to diagnose the aetiology of The IPSS index would probably perform less well whendiscriminating unselected patients of similar ages withthe symptoms. This does not necessarily compromise the

40 B.J. HANSE N et al.

and without LUTS suggestive of BOO. In the IPSS might be enough to lessen discomfort and worry,diminish bother and improve functioning, i.e. a slightvalidation process, Barry et al. [6] adapted the Madsen-

Iversen score system for self-administration, with inevi- improvement will not be detected clearly in the IPSS butwill in the DAN-PSS index. To measure symptomatictable modifications which may have invalidated the

correlation. The Madsen-Iversen system might have per- outcome, a questionnaire must include items related tomore important complications or untoward results. Theformed diCerently if the questions were asked and point

scores determined by an interviewer. The correlation DAN-PSS has four questions related to incontinence,whereas the IPSS index includes no questions related tocoeBcient was 0.85 [6] but in the present study, where

the Madsen-Iversen score was determined by interview, incontinence.The DAN-PSS has excellent reliability and validity andthe correlation between it and the IPSS was 0.45

(Fig. 1c), a poorer correlation than that determined has been tested extensively; it is more sensitive thanthe IPSS index, incorporates important outcome events,previously.

The correlation coeBcient between the DAN-PSS and includes a patient-weighting of each symptom, therebyreflecting better the patient’s global assessment ofIPSS indexes was 0.61 (Fig. 1d); this weak correlation

showed that there was some overlap in the concepts outcome.As quality of life concerns become more predominantincluded in the two scoring systems. Fewer questions are

included in the IPSS and the questions contained are in the treatment of LUTS suggestive of BOO, and otherdisorders, the use of symptom-scoring systems willmainly concerned with the temporal occurrence of symp-

toms included and not the severity, as is the case for expand and their degree of sophistication continue toincrease. The ICS is currently evaluating a questionnairequestions in the DAN-PSS system. Consequently, more

information about the condition might be obtained from similar to the DAN-PSS [20,21]. The objectives of theICS-BPH study are to develop a valid and reliable tool tothe latter.

The DAN-PSS is more responsive than the Boyarsky measure the prevalence and severity of LUTS and theireCects on quality of life, and to investigate the relation-system after TUMT (Fig. 3). There was a reduction in

the DAN-PSS index of 57%, in contrast to that in the ships between the ICS questionnaire and the results ofadvanced urodynamics. Such an undertaking will helpBoyarsky index of 15%, after 12 months. The Boyarsky

system weights items related to voiding function more to clarify the extent to which urodynamics should beincluded in the evaluation of LUTS suggestive of BOO.than items related to storage function; moreover, the

Boyarsky index lacks a bother factor. The DAN-PSS sumsthe products of each individual symptom and its ‘bother’

Referencesgrading. The incorporation of a ‘bother factor’ in theDAN-PSS questionnaire for each question may explain 1 Mebust WK, Holtgrewe HL, Cockett ATK, Peters PC.

Transurethral prostatectomy: immediate and postoperativethis diCerence. The DAN-PSS system is also significantlycomplications. A cooperative study of 13 participatingmore responsive than the IPSS (Fig. 2). There are severalinstitutions evaluating, 3885 patients. J Urol 1989;possible interpretations of the evidence for the lower141: 243–7responsiveness of the IPSS index. Using the IPSS index

2 Mebust W, Roizo R, Schroeder F, Villiers A. Correlationinvolves the risk of treating symptomatic patients whobetween pathology. clinical symptoms and the course of

are not bothered by their symptoms. Furthermore, thethe disease. In: Cockett ATK, Aso Y, Chatelain C et al.

IPSS system omits questions related to various types of eds, The International Consultation on Benign Prostaticincontinence, symptoms known to be of major concern Hyperplasia (BPH). SCI, Geneva, WHO, 1991: 53–62to the patients, and weights voiding more than storage 3 Boyarsky S, Jones G, Paulson DF, Prout GR Jr. A newsymptoms, which contradicts patients’ perceptions [18] look at bladder neck obstruction by food administration

regulators: guidelines for the investigation of benign(Table 1). In addition, the IPSS system lacks a botherprostatic hyperplasia. Trans Am Ass Genitourin Surg 1977;factor for each question. Emberton et al. [19] showed in68: 29–32a study of 5276 men undergoing prostatectomy that

4 Madsen PO, Iversen P. A point system for selectingchanges in symptom severity were highly correlatedoperative candidates. In Hinman F Jr, ed., Benign Prostaticwith changes in bothersomeness and disease-specificHypertrophy. New York: Springer, 1983: 763–5quality of life. Each symptom in the DAN-PSS question-

5 Barry M, Fowler FJ Jr, O’Leary MP et al. and thenaire is weighted by the patient’s bother of the symptom

measurement committee of the American Urologicaland the impact on his daily living (bother factor). Thus, Association. The American Urological Associationa symptom can be disregarded if the patient is not Symptom Index for benign prostatic hyperplasia. J Uroltroubled by it and a slight symptom causing severe 1992; 148: 1549–57bother will be weighted accordingly higher. A slight 6 Barry M, Fowler FJ Jr, O’Leary MP et al. and the

measurement committee of the American Urologicalimprovement rating by patients in the symptom score

COMPARISON OF SYMPTOM SCORES 41

Association. Correlation of the American Urological prostatectomy on symptom severity and quality of life. BrJ Urol 1996; 77: 233–47Association symptom index with the self-administered

version of the Madsen-Iversen, Boyarsky and Maine Medical 20 Donovan JL, Abrams P, Peters TJ et al. The ICS-BPH study:the psychometric validity and reliability of the ICSmaleAssessment programme symptom indexes. J Urol 1992;

148: 1558–63 questionnaire. Br J Urol 1996; 77: 554–6221 Peters J, Donovan JL, Kay HE et al. and the International7 Hansen BJ, Flyger H, Brasso K et al. Validation of the self-

administred Danish Prostatic Symptom Score (DAN-PSS-1) Continence Society ‘Benign Prostatic Hyperplasia’ studygroup. The International Society ‘Benign Prostaticsystem for use in benign prostatic hyperplasia. Br J Urol

1995; 76: 451–8 Hyperplasia’ study: the bothersomeness of urinarysymptoms. J Urol 1997; 157: 885–98 Hansen BJ, Nordling J, Mensink HJ et al. and the Alfech

study group. Alfuzosin in the treatment of benign prostatichyperplasia: ECects on symptoms, urinary flow rates andresidual volume. A multicentre, double-blind, placebo-controlled trial. Scand J Urol Nephrol 1994; 157 (suppl): Authors169–76

B.J. Hansen, MD, PhD, Senior Registrar.9 Hendolin N, Bolmsjo M, Alfthan O. Transurethral micro-

S. Mortensen, MD, DMSc, Chief Urologist.wave thermotherapy in patients with symptomatic prosta-

H.J.A. Mensink, MD, DMSc, Professor.tism due to benign prostatic hyperplasia or chronic

H. Flyger, MD, PhD, Registrar.prostatitis. Scand J Urol Nephrol 1995; 169 (suppl): 41–2

M. Riehmann, MD, Registrar.(abstract 50)

N. Hendolin, MD, Registrar.10 Kirshner B, Guyatt G. A methodological framework for

J. Nordling, MD, DMSc, Chief Urologist.assessing health indices. J Chron Dis 1985; 38: 27–36

T. Hald MD, DMSc, Professor, Chief Urologist.11 Hansen BJ, Flyger H, Mortensen S et al. and the Alfech

Correspondence: Dr B.J. Hansen, Julmosevej 16, DK-3460study group. Symptomatic outcome of transurethral prosta-

Birkerod, Denmark.tectomy, alpha-blockade and placebo in the treatment ofbenign prostatic hyperplasia. Scand J Urol Nephrol 1996;30: 103–7 Appendix 1

12 Lepor H, Rigaud G. The eBcacy of transurethral resectionThe Danish–Dutch cooperative ALFECH study group.of the prostate in men with moderate symptoms ofThe Netherlands:prostatism. J Urol 1990; 143: 533–7

13 Lepor H, Williford WO, Barry MJ et al. for the Veterans JAF Leenarts, Westeinde Hospital, LijnbaanACairs cooperative studies benign prostatic hyperplasia JK Oosten, Ikazia Hospital, Rotterdamstudy group. The eBcacy of terazosin, finasteride, or both FF van Soest, Hospital Gooi-Noord, Blaricumin benign prostatic hyperplasia. N Engl J Med 1996; GA Dijkman, St Ignatius Hospital, Breda335: 533–9 JW Hoekstra, Bosch Medicentr, Den Bosch

14 Lepor H, Machi G. Comparison of the AUA symptom index NJW van Baasbank, Hospital De Weezenlanden, Zwollein unselected males and females between fifty-five and

RTh Bijleveld, St Antonius Hospital, Nieuwegeinseventy-nine years of age. Urology 1993; 42: 36–41

PFCM Braam, St Joseph Hospital, Veldhoven15 Chai TC, Belville WD, McGuire EJ, Nyquist L. Specificity ofThJM Schlatmann, Olug, Amsterdamthe American Urological Association symptom index:J Felderhof, Spaarne Hospital, Haarlemcomparison of unselected samples of both sexes. J UrolBJ Kapper, Medisch Spectrum Twente, Enschede1993; 150: 1710–3

16 Yalla SV, Sullivan MP, Lecamwasam HS, Dubeau CE, P Dik, Centraal Militair Hospital, UtrechtVickers MA, Cravalho EG. Correlation of American HJA Mensink, Academic Hospital, GroningenUrological Association symptom index with obstructive Denmark:and nonobstructive prostatism. J Urol 1995; 153: 674–80 J Nordling, J Schou, AL Poulsen, KAS Herlev, Herlev

17 Poulsen AL, Schou J, Puggard L, Torp-Pedersen S, J ChristoCersen, Amtssygehuset in Fakse, FakseNordling J. Prostatic enlargement, symptomatology and JP Geerdsen, RASK, Køgepressure/flow evaluation: interrelations in patients with

V Hvidt, C Dahl, M Luke, BJ Hansen, Bispebjerg Hospital,symptomatic BPH. Scand J Urol Nephrol 1994; 157

København NV(suppl): 67–73HH MeyhoC, A Lendorph, B Jacobsen, Centralsygehuset18 Jolleys JV, Donovan JL, Nanchahal K, Peters TJ, Abrams P.

in Hillerød, HillerødUrinary symptoms in the community: how bothersome areT Bilde, S Mortensen, KAS Glostrup, Glostrupthey? Br J Urol 1994; 74: 551–5

19 Emberton M, Neal DE, Black N et al. The eCect of S Walter, EH Larsen, Aalborg Sygehus Nord, Aalborg