Facebook Twitter YouTube

18571

Objective Measurement of Voiding Function Among Vesicoureteral Reflux Patients

Saturday, October 20, 2012
Grand Ballroom A/B (Hilton Riverside)
Katherine W. Herbst, M.Sc., John H. Makari, MD, FAAP, Fernando A. Ferrer Jr., MD, FAAP and Christina Kim, MD, Division of Urology, Connecticut Children's Medical Center, Hartford, CT

Purpose:

Dysfunctional Elimination Syndrome (DES) is a prevalent abnormality in patients with vesicoureteral reflux (VUR). Objective measures may not be predictive of disease presence. The purpose of this study was to assess the accuracy of two objective measures, the Dysfunctional Voiding Symptom Score (DVSS) and post-void residual as a percent of estimated bladder capacity (PVR/EBC) when compared to the clinician's diagnosis of DES.

 

Methods:

In June 2010, we created a prospective registry for anti-VUR surgery patients.  Since then, fifty-one patients have been enrolled, forty-six have undergone surgery, and forty-four have had follow up visits. Clinician diagnosis of DES (DxDES) was obtained before surgery and at follow-up visits by two clinicians. Pre-surgical DVSS score was recorded for 43/51 patients, and follow-up DVSS for 36/44 patients. Pre-surgical PVR /EBC was recorded for 42/51 patients, and follow-up PVR/EBC for 38/44 patients. Patients were scored as positive for DES if DVSS score was >6 for females and >9 for males. To determine the precision of DVSS, DxDES was used as the gold standard and positive predictive value (PPV), sensitivity and specificity calculated. A receiver operating characteristic (ROC) curve was used to determine the accuracy of PVR/EBC with DxDES as the gold standard.

 

Results:

The majority of patients had a DxDES before surgery (57%), with only 25% showing DxDES after surgery. Pre-surgical DVSS had a PPV of 75%, a sensitivity of 56%, and a specificity of 69% (Table 1). Post-surgical DVSS had a PPV of 40%, a sensitivity of 36%, and a specificity of 24% (Table 2). ROC curves for both pre-surgical and follow-up PVR/EBC had area under the curve values of 0.456 and 0.464 respectively (Figures 1 & 2).

 

 

Conclusions:

Although pre-surgical DVSS was somewhat precise in identifying DxDES, sensitivity and specificity were low. Post-surgical DVSS statistics were very inexact. The PVR/EBC ratio (at any level) was unable to correctly identify a clinician's diagnosis of DES with any greater frequency than a random guess as determined by ROC curve analysis. Given these results, we theorize that our clinicians rely on a multitude of factors other than DVSS and PVR.  We recognize this study is limited by sample size and non-generalizability.  Nonetheless, these findings raise questions regarding the value of the DVSS and PVR/EBC as diagnostic tests to assess DES.  We believe further investigation of objective measures is warranted.

Table 1. Pre-surgical Findings

                                                  DxDES

                                    +                                  -                                  

                        +          15                                5                      PPV

DVSS                                                     Type I Error             0.75    

                        -           12                                11                    NPV

                               Type II Error                                            0.48    

                              Sensitivity                 Specificity

                                    0.56                         0.69

 Table 2. Follow-up Findings

                                                  DxDES

                                    +                                  -                                  

                        +          4                                6                      PPV

DVSS                                                     Type I Error             0.40    

                        -           7                                19                    NPV

                               Type II Error                                            0.73    

                              Sensitivity                 Specificity

                                    0.36                         0.24