AIMS OF STUDY
Patients with severe Genuine Stress Incontinence (GSI), a fixed urethra and no hypermobility, are ideal candidates for studying Intrinsic Sphincter Deficiency (ISD) as opposed to anatomic incontinence. We analyzed clinical data (age, previous uro-gynaecological surgery and/or hysterectomy, stress test, irritative symptoms and perineal testig) and urodynamic parameters (Valsava leak point pressure - VLPP, Maximum Urethral closure pressure - MUCP, the urethral functional length), in a group of patients with severe GSI and poor urethral mobility (pure ISD) and with moderate GSI and urethral hypermobility (pure anatomic GSI) in an attempt to identify specific ISD patterns (1-2-3),

METHODS
We recruited 87 consecutive patients with urodynamically demonstrated GSI. Each woman provided a standard urogynecologic based history, and underwent physical examination and full urodynamic testing. The VLPP and the MUCP were determined with 200 ml bladder volume, using an infusion pump and an 8 Fr catheter. VLPP was determined during a Valsalva manouvre. A MUCP £30 cm H20 and a VLPP £ 60 cm H20 were chosen as cut?off values. Poor or no urethral mobility was diagnosed with urethrocele £1 (HWS classification) associated with ultrasound cervicourethral displacement of £ 1 cm. Severity of incontinence was subjective (SEAPI?QMN classification). 38/87 pts had incontinence grade ³ 2 and urethrocele £ 1 (pure ISD), 49/87 pts had incontinence grade = 1 and urethrocele ³2 (pure anatomic GSI - no ISD) Statistical analysis.For the logistic regression model, the disease status was used as binary dependent variable (48 pts with no ISD: disease status =0; 39 pts with pure ISD: disease status =1) and was correlated against these independent variables: age, previous uro-gynaecological surgery and/or hysterectomy, stress test, irritative symptoms, perineal testing, MUCP, VLPP and urethral functional length. The odds of having ISD were defined from odds ratio (OR) coefficients and 95% confidence interval (CI) using logistic regression. To determine the odds ratios, a reduced model of stepwise logistic regression was produced, incorporating the same predictors, by backward conditional elimination that included only statistically significant potentially predictive variables. All statistical procedure were carried out in SPSS for Windows.

RESULTS
The multiple logistic analysis showed that only a low VLPP (below the cut off value) (r=0,314; p<0.0001), previous urogynecologycal surgery and/or hysterectomy (r=0.294; p<0.001) have a significant correlation with the presence of ISD. The correct estimate of this model was 81.9% for total subjects, 84.4% for subjects with disease status = 0 and 80% for subjects with disease status = 1. A stepwise logistic regression was used incorporating the VLPP value and the presence or absence of previous surgery. The low VLPP determines an odds ratio = 1.9 (CI 1.1-3); previous surgery determines an odds ratio = 2.1 (CI 13-3.4). When a low VLPP is associated with previous surgery the odds ratio is = 19 (7.1-50.7).

CONCLUSIONS
Logistical analysis of data from this series of patients shows only the VLPP and previous urogynaecological surgery and/or hysterectomy correlated with ISD. Low VLPP identifies 65% of cases while previous surgery detects 59%. When both variables are associated 89% of patients with ISD are identified.

REFERENCES
1. MeGuire EJ, Fitzpatrick CC, Wan J, Bloorn D, Sanvordenker J, Ritchey M Gormley EA: Clinical assessment of urethral sphincter function. J Urol., 150: 1452. 1993
2. Sand PK, Bowen LW, Panganiban P, Ostergard DR: The low pressure urethra as a factor in failed retropubic urethropexy. Obstet Gynecol., 69:399. 1987 3. C. Pajoncini, C. Radicchia, E. Costantini, R. Lombi, F. Guercini, L. Mearini, M. Porena.: Leak point pressure in the diagnosis of intrinsic sphinteric deficiency. Neurourol. Urodyn., 17:389-391. 1998