1. Top of page
  2. Introduction
  3. Clinical utility of PFS
  4. Tests substituting for PFS
  5. Treatment schemes and expected outcomes
  6. Conclusions
  7. References

There is controversy about whether pressure-flow studies (PFS) are mandatory to indicate surgical treatment for men with BPH [1,2]. Those who advocate PFS insist that ‘obstructed’ should be discriminated from ‘unobstructed’ men to improve treatment outcome, and that there seems to be an ‘a priori’ agreement that PFS should be undertaken before prostatectomy [3]. To the contrary, those who do not mandate PFS argue that the precise distinction between ‘obstructed’ and ‘unobstructed’ does not improve the surgical outcome sufficiently to justify the cost and invasiveness of the procedure [4]. In other words, surgical outcomes have been satisfactory without PFS before surgery [5], or a careful assessment of conventional study results has provided sufficient information to proceed with treatment. However, these authors agree that PFS represent the ‘gold standard’ for establishing the urodynamic diagnosis of obstruction, and should be used in investigational clinical trials or when other underlying causes for the voiding dysfunction may prevail [1–4]. In this short review the impact of PFS test results on surgical outcomes is discussed, by proposing various treatment schemes for BPH.

Clinical utility of PFS

  1. Top of page
  2. Introduction
  3. Clinical utility of PFS
  4. Tests substituting for PFS
  5. Treatment schemes and expected outcomes
  6. Conclusions
  7. References

The utility of a clinical test should be assessed by the balance between the information obtained through the test and the cost required. In the present case, the information obtained is the precise quantification of bladder outlet obstruction (BOO). The cost includes the purchase and maintenance of equipment, medical personnel resources consumed, or time and money paid by patients.

The cost is so multifactorial, complex and variable among countries that an element of cost (invasiveness, i.e. the discomfort and adverse events associated with the procedure) is sometimes placed as the counterbalance to the information. However, despite this simplification, invasiveness remains obscure; it can be defined only in relative terms. Compared with uroflowmetry or transabdominal ultrasonography, PFS are obviously more invasive, but how invasive is TRUS or cystoscopy? Moreover, the degree of invasiveness associated with PFS would be subject to large variability among examiners and individual patients, which would necessitate individualized estimates of invasiveness.

The utility of information on the presence or absence of obstruction can be appreciated as the difference in surgical outcomes (Table 1, [6–8]). Although patient backgrounds or response criteria were not consistent among the studies, there were advantages in success rate by 15–29% for obstructed patients. Nevertheless, obstructed men did not always fare well (success rate 79–93%) and conversely, unobstructed men did not always fail, with moderate success rates of 55–78%. The difference is smaller for the long-term outcome at 8 years, i.e. an 83% success rate in obstructed and 72% in unobstructed men [9]. Immature surgical techniques and/or persistent detrusor instability [6] may account for the failure in obstructed patients. Success in unobstructed men might be attributable to surgical effects other than relieving obstruction, e.g. placebo effects; it could be also explained by an inherent limitation of PFS in detecting obstruction, when obstruction and a weak detrusor coexist [10], i.e. low detrusor pressure does not necessarily contraindicate prostatectomy [11].

Table 1.  Surgical success rate and BOO
StudySurgical success rate, % obstructed/unobstructedDifference, %

Tests substituting for PFS

  1. Top of page
  2. Introduction
  3. Clinical utility of PFS
  4. Tests substituting for PFS
  5. Treatment schemes and expected outcomes
  6. Conclusions
  7. References

Although the extent cannot be clearly defined, PFS are more invasive than most other clinical investigations that are potentially used for determining obstruction. Therefore it is logical to explore the possibility of replacing PFS by one or a combination of these tests for diagnosing obstruction. To date, several clinical tests have been assessed for predicting obstruction (Table 2, [12–20]). Test items have included the postvoid residual volume (PVR), maximum urinary flow rate (Qmax), prostate volume, endoscopic findings suggestive of obstruction, estimated bladder weight, presumed roundness of the prostate on ultrasonography, or derived values calculated from regression analyses. Symptoms have repeatedly been shown to be a poor predictor of obstruction [19–22] and are thus excluded from Table 2.

Table 2.  The capability of various tests for predicting BOO [12–20]
StudyTest itemObstructed /unobstructedPPVNPVSensitivitySpecificity
  • *

    Estimated from graphical presentation; NE, not examined; PPV, positive predictive value; NPV, negative predictive value.

[12]PVR≥ 50/ < 50 mL86256948
[13]Qmax≤ 10/ > 10 mL/s75506957
Qmax≤ 15/ > 15 mL/s71639031
Prostate volume> 50/ ≤ 50 mL78363480
[14]CystoscopyGrade 3/Grade 1,290*30*NENE
[15]BOO number≥ −2/ < −29037*6674
[16]Prostate score≥ 11/ < 11NENE8153
[17]Bladder weight> 35/ ≤ 35 g88848587
[18]Prostate roundness≥ 0.8/ < 0.886627775
[19]Qmax< 15/ ≥ 15 mL/s88429717
Prostate volume≥ 30/ < 30 mL77235290
UrethroscopyGrade 2/Grade 0,197196091
[20]Qmax< 15/ ≥ 15 mL/s67428238

Interestingly, with the thresholds used in these studies, positive predictive values (PPVs) are larger than negative predictive values (NPVs). The PPV represents the proportion of obstructed men in those testing positive, and thus it should be higher than the NPV because these tests are intended to select obstructed men. The sensitivity is larger than the specificity when using the PVR or Qmax, but vice versa with endoscopy or prostate volume. What differs among these tests? Tests with high sensitivity and low specificity can efficiently detect obstructed men but tend to falsely diagnose unobstructed men as obstructed (high false-positive ratio). Tests with low sensitivity and high specificity are more likely to miss obstructed men (high false-negative ratio), but can prudently exclude unobstructed men. Thus a common protocol is to use PVR or Qmax as the discriminator, as they have high sensitivity and low specificity, and can effectively identify obstructed men but may retain a substantial number of ‘falsely’ obstructed men. A combination of items with distinct test characteristics might provide a more stringent exclusion of unobstructed men with less chance of missing obstructed men [15,16,19,21].

Treatment schemes and expected outcomes

  1. Top of page
  2. Introduction
  3. Clinical utility of PFS
  4. Tests substituting for PFS
  5. Treatment schemes and expected outcomes
  6. Conclusions
  7. References

For simplicity, patient outcomes are modelled for two hypothetical treatment schemes (A and B). In scheme A, all the patients undergo TURP with no evaluation by PFS. In contrast, in scheme B, all patients are examined by PFS and only those with confirmed obstruction undergo TURP. The diagnostic accuracy of PFS for obstruction is assumed to be 100%. The success rate of TURP in scheme B (Sb) is identical to that for obstructed patients (So). To calculate the success rate of TURP in scheme A (Sa), the ratio of obstructed patients to the total (p) is postulated. Then

  • image

where Su is the success rate of TURP for unobstructed patients. Thus

  • image

Figure 1 shows the relationship between (Sb − Sa) and (So − Su); the hatched area represents a realistic value of (So − Su) [6–8] and p [23]; thus (Sb − Sa) would be 2–12% and such a small difference would explain the impression that surgical outcomes have been satisfactory with no PFS before TURP [4,5].


Figure 1. Two treatment schemes are assumed; all patients undergo TURP (scheme A), and all patients are examined by PFS first, with only obstructed men undergoing TURP (scheme B). The plot shows the relationship between (Sb − Sa) and (So − Su) with p calculated as given in the text; the red hatched area represents a realistic value of (So − Su) (10–30%) and of p (0.6–0.8), with (Sb − Sa) at 2–12%.

Download figure to PowerPoint

A more complicated assumption with a specific number of patients is given in Table 3. Suppose there is a population of 100 patients with BPH who are awaiting TURP; the assumptions include: (i) 70 of them are obstructed and the remaining 30 unobstructed; (ii) the success rate of TURP for obstructed men is 90% and that for unobstructed is 60%; and (iii) the sensitivity and specificity for predicting obstruction is 100% in PFS, whilst these values are variable for other diagnostic modalities (assumed later).

Table 3.  The treatment schemes and patient outcome. The model population is 100 patients with BPH comprising 70 obstructed men and 30 unobstructed men. The surgical success rate is assumed to be 90% for obstructed and 60% for unobstructed men. In scheme A all men undergo TURP with no clinical tests or PFS. In scheme B, all men undergo PFS (which have 100% accuracy) and only obstructed men undergo TURP. In scheme C1, all men first undergo various tests with a sensitivity of 80% and specificity of 40%, and only men considered obstructed on these tests undergo TURP. In scheme C2, unobstructed men in scheme C1 (not undergoing TURP) undergo further PFS and obstructed men on PFS undergo TURP. The detailed algorithm is not shown for schemes D and E (see text).
Success/failure for
SchemeObstructed TURP/ no TURPNot obstructed TURP/ no TURPObstructedNot obstructedTURP

The patients are treated according to one of eight treatment schemes; schemes A and B are the same as in Fig. 1. In scheme A, all patients undergo TURP with no further clinical tests and in scheme B the patients are all examined with PFS first, and only those with obstructed BPH undergo TURP. In schemes C1, D1 and E1, the patients have their bladder outlet examined by various clinical tests other than PFS and only those diagnosed as obstructed by these tests undergo TURP. Finally, in schemes C2, D2 and E2, the patients designated as unobstructed in schemes C1, D1 and E1 are further investigated with PFS, and those with confirmed obstruction by PFS also undergo TURP. The sensitivity and specificity of the other tests are assumed to be 80% and 40% in scheme C, 40% and 80% in scheme D, and 80% and 80% in scheme E, respectively.

The patient outcomes of these schemes are summarized in Table 4. The overall success rate is best in scheme A (81%), followed by the group of schemes C2, D2 and E2 (79–77%), scheme B (75%) and the group of schemes C1, D1 and E1 (67–43%). The low overall success rates in schemes C1, D1 and E1 (67–43%) suggest that a surgical indication based solely on clinical tests, except PFS, may result in worse overall outcomes than no selective indication (scheme A). This is particularly true for tests with a low sensitivity (scheme D1, 43%). The paradox occurs because most patients with BPH are obstructed, and TURP is considered very effective for obstructed men and moderately effective for unobstructed men. Thus the simple indication for TURP with no specific considerations would offer a fairly good outcome (scheme A). However, the treatment policy in other schemes is that only obstructed men undergo TURP, although the identification of obstruction is not perfect. Consequently, undertaking these tests prevents some obstructed men from receiving the effective treatment, i.e. some obstructed men are wrongly diagnosed by the tests as unobstructed.

Table 4.  A summary of outcomes of 100 patients following the treatment schemes in Table 3
SchemeNo. undergoingOverall outcomeSurgical outcome, n (%)Decrease in surgical failure
PFSSurgerySuccessFailureSuccessFailure19 − no. of failures*No. of PFS/decrease
  • *

    Number of decreases in surgical failure from the worst (scheme A, 19);

  • †Number of PFS needed to produce one decrease in surgical failure from the worst (scheme A, 19). See text for the detailed treatment schemes. The sensitivity and specificity of various tests are postulated as 80% and 40% in scheme C, 40% and 80% in scheme D, and 80% and 80% in scheme E, respectively.

A0100811981 (81)19 (19)0
B10070752563 (90)7 (10)128.3 (100/12)
C1074673361 (82)13 (18)6
C22688792174 (84)14 (16)55.2 (26/5)
D1034435729 (85)5 (15)14
D26676772367 (88)9 (12)106.6 (66/10)
E1062653554 (87)8 (13)11
E23876772367 (88)9 (12)103.8 (38/10)

These outcomes are much improved if PFS are also used to discriminate obstructed from ‘falsely’ unobstructed men (schemes C2, D2 and E2). The success rates in these latter schemes are reasonably high (77–79%), irrespective of the sensitivity and specificity of the first tests. The surgical success rate is distributed within a narrow range of 9%, between the best in scheme B (90%) and the worst in scheme A (81%). However, focusing on the number of surgical failures, the worst (19, scheme A) is 2.7 times more than the failure in scheme B (seven). Therefore, clinical tests including PFS may not substantially increase the patients' benefit from surgery. By contrast, these tests can reduce the harm of useless surgery, and indeed the utility of these tests should be judged on the basis of preventing failure rather than enhancing success.

Preventing surgical failure can be evaluated as the difference in the number of failures from the worst value (19, scheme A; Table 4). The difference is greatest in scheme D1 (14), because the test with high specificity can effectively exclude unobstructed men, affording a more conservative and safer indication. The efficacy of incorporating PFS into pretreatment tests can be evaluated as the number of PFS that are needed to decrease one surgical failure from the worst failure in scheme A (19). The value is largest in scheme B (8.3); the strategy ‘PFS in all subjects’ is less effective than ‘PFS in selected subjects’.

Again, suppose there are 100 patients with BPH; assume a sensitivity for uroflowmetry of 80% and a specificity of 40% (scheme C), and for prostate volume a sensitivity of 40% and a specificity of 80% (scheme D). The simplest protocol is that all undergo TURP with no further evaluation (scheme A); it has a satisfactory outcome but gives the most surgical failures (19 of 100 patients). Evaluating all patients by PFS is the most labour-intensive diagnostic procedure (scheme B), but it has the fewest surgical failures (seven); results from the other schemes lie between these. For example, when if uroflowmetry is initially used for the 100 men, then PFS in the 26 men with a Qmax of more than the threshold (scheme C2), there are 14 surgical failures. When the prostate volume is measured first, followed by PFS in 66 men with a small prostate (scheme D2), there are nine surgical failures. If there was an almost ideal test with 80% sensitivity and 80% specificity (scheme E2), only 38 men need to undergo PFS to achieve the same results.


  1. Top of page
  2. Introduction
  3. Clinical utility of PFS
  4. Tests substituting for PFS
  5. Treatment schemes and expected outcomes
  6. Conclusions
  7. References

This examination of various treatment schemes, albeit based on many assumptions, provides an idea of the extent to which surgical failure is prevented by using PFS and other clinical tests. The degree of prevention should be balanced with that of the invasiveness associated with the tests which, as noted earlier, is estimated individually. Hypothetical but straightforward information about perceived outcomes in these schemes may be useful in obtaining informed consent for diagnostic and therapeutic alternatives for suspected BOO [24].


  1. Top of page
  2. Introduction
  3. Clinical utility of PFS
  4. Tests substituting for PFS
  5. Treatment schemes and expected outcomes
  6. Conclusions
  7. References
  • 1
    Abrams P, Buzelin JM, Griffiths D et al. The urodynamic assessment of lower urinary tract symptoms. In CockettATK, KhouryS, AsoY et al. eds, International Consultation on Benign Prostatic Hyperplasia. 4th edn. Paris: Scientific Communication International, 1998: 32378
  • 2
    Gerber GS. The role of urodynamic study in the evaluation and management of men with lower urinary tract symptoms secondary to benign prostatic hyperplasia. Urology 1996; 48: 66875
  • 3
    Abrams P. In support of pressure-flow studies for evaluating men with lower urinary tract symptoms. Urology 1994; 44: 1535
  • 4
    McConnell JD. Why pressure-flow studies should be optional and not mandatory studies for evaluating men with benign prostatic hyperplasia. Urology 1994; 44: 1568
  • 5
    Kalpan SA, Bowers KS, Te AE, Olsson CA. Differential diagnosis of prostatism: a 12-year retrospective analysis of symptoms, urodynamics and satisfaction with therapy. J Urol 1996; 155: 13038
  • 6
    Jensen KME, Jorgen JB, Mogensen P. Urodynamics in prostatism; II. Prognostic value of pressure-flow study combined with stop-flow test. Scand J Urol Nephrol Supple 1988; 114: 727
  • 7
    Neal DE, Ramsden PD, Sharples L et al. Outcome of elective prostatectomy. Br Med J 1989; 299: 7627
  • 8
    Robertson AS, Griffiths C, Neal DE. Conventional urodynamics and ambulatory monitoring in the definition and management of bladder outflow obstruction. J Urol 1996; 155: 50611
  • 9
    Jensen KME, Jorgensen JB, Mogensen P. Long-term predictive role of urodynamics: an 8-year follow-up of prostatic surgery for lower urinary tract symptoms. Br J Urol 1996; 78: 2138
  • 10
    Gotoh M, Yoshikawa Y, Kondo AS, Kondo A, Ono Y, Ohshima S. Prognostic value of pressure-flow study in surgical treatment of benign prostatic obstruction. World J Urol 1999; 17: 2748DOI: 10.1007/s003450050145
  • 11
    Ball AJ & Smith PJB. Urodynamic factors in relation to outcome of prostatectomy. Urology 1986; 28: 2568
  • 12
    Abrams P & Griffiths DJ. The assessment of prostatic obstruction from urodynamic measurements and from residual urine. Br J Urol 1979; 51: 12934
  • 13
    Poulsen A, Schou J, Puggaard L, Torp-Pedersen S, Nordling J. Prostatic enlargement, symptomatology and pressure/flow evaluation: Interrelations in patients with symptomatic BPH. Scand J Urol Nephrol Suppl 1994; 157: 6773
  • 14
    Ezz EI, Din K, De Wildt MJAM, Rosier PFWM, Debruyne FFMJ, De La Rossette JJMCH. The correlation between urodynamic and cystoscopic findings in elderly men with voiding complaints. J Urol 1996; 155: 101822
  • 15
    Van Venrooij GEPM & Boon TA. The value of symptom score, quality of life score, maximal urinary flow rate, residual volume and prostate size for the diagnosis of obstructive benign prostatic hyperplasia: a urodynamic analysis. J Urol 1996; 155: 20148
  • 16
    Roiser PFWM, De Wildt MJAM, Wijkstra H, Debruyne FFMJ, De La Rossette JJMCH. Clinical diagnosis of bladder outlet obstruction in patients with benign prostatic enlargement and lower urinary tract symptoms: development and urodynamic validation of a clinical prostate score for the objective diagnosis of bladder outlet obstruction. J Urol 1996; 155: 164954
  • 17
    Kojima M, Inui E, Ochiai A, Naya Y, Ukimura O, Watanabe H. Noninvasive quantitative estimation of infravesical obstruction using ultrasonic measurement of bladder weight. J Urol 1997; 157: 4769
  • 18
    Kojima M, Ochiai A, Naya Y, Ukimura O, Watanabe M, Watanabe H. Correlation of presumed circle area ratio (PCAR) with infravesical obstruction in men with lower urinary tract symptoms. Urology 1997; 50: 54855
  • 19
    Homma Y, Gotoh M, Takei M, Kawabe K, Yamaguchi T. Predictability of conventional tests for bladder outlet obstruction assessment in benign prostatic hyperplasia. Int J Urol 1998; 5: 616
  • 20
    Reynard JM, Yang Q, Donovan JL et al. The ICS-‘BPH’ Study: uroflowmetry, lower urinary tract symptoms and bladder outlet obstruction. Br J Urol 1998; 82: 61923
    Direct Link:
  • 21
    Madersbacher S, Klingler HC, Djavan B et al. Is obstruction predictable by clinical evaluation in patients with lower urinary tract symptoms? Br J Urol 1997; 80: 727
  • 22
    Yalla SV, Sullivan MP, Lecamwasam HS, DuBeau CE, Vickers MA, Cravalho EG. Correlation of American urological association symptom index with obstructive and nonobstructive prostatism. J Urol 1995; 153: 67480
  • 23
    Nielsen KK, Nordling J, Hald T. Critical review of the diagnosis of prostatic obstruction. Neurourol Urodyn 1994; 13: 20117
  • 24
    Arnold EP. Medicolegal cases relating to bladder outlet obstruction. BJU Int 2000; 86: 30812DOI: 10.1046/j.1464-410x.2000.00299.x