ABILITY OF A NORMAL DYSFUNCTIONAL VOIDING SYMPTOM SCORE TO PREDICT UROFLOWMETRY AND EXTERNAL URINARY SPHINCTER ELECTROMYOGRAPHY PATTERNS IN CHILDREN

0022-5347/04/1725-1980/0 THE JOURNAL OF UROLOGY® Copyright © 2004 by AMERICAN UROLOGICAL ASSOCIATION

Vol. 172, 1980 –1985, November 2004 Printed in U.S.A.

DOI: 10.1097/01.ju.0000140849.49348.62

ABILITY OF A NORMAL DYSFUNCTIONAL VOIDING SYMPTOM SCORE TO PREDICT UROFLOWMETRY AND EXTERNAL URINARY SPHINCTER ELECTROMYOGRAPHY PATTERNS IN CHILDREN DONALD P. BARTKOWSKI

AND

RUSSELL G. DOUBRAVA

From the Michigan State University College of Osteopathic Medicine, Sparrow Health Systems, East Lansing, Michigan

ABSTRACT

Purpose: We correlated uroflowmetry/electromyography (EMG) patterns and post-void residual urine volume (PVR) in children with a normal dysfunctional voiding symptom score (DVSS) and determined whether a normal DVSS can predict uroflowmetry/EMG/PVR results. Materials and Methods: We prospectively studied 51 children with a normal DVSS, and negative neurological and urological history who underwent uroflowmetry/EMG and PVR measurements. Statistical results were correlated with patient age and gender using Cohen’s ␬ and Fisher’s exact test. Results: All children had a normal DVSS for gender and 37 (73%) had bell-shaped uroflowmetry curves. Eight of the 14 children with nonbell-shaped uroflowmetry curves had plateaued and 6 had fractionated curves. Five boys and 13 girls (35%) had decreased EMG patterns, 7 boys and 8 girls (29%) had increased patterns and 5 boys and 13 girls (35%) had variable patterns (Cohen’s ␬ score 0.11). Of the 37 children (65%) with bell-shaped uroflowmetry curves 24 had normal PVR, while 13 of 37 (35%) had elevated PVR. Eight of the 14 children (57%) with nonbell-shaped uroflowmetry curves had normal PVR and 6 (43%) had elevated PVR (Cohen’s ␬ score 0.13). Seven of 37 children (19%) with bell-shaped uroflowmetry curves and 7 of 14 (50%) with nonbell-shaped uroflowmetry curves had a significantly distended bladder (p ⫽ 0.04). Nine of the 14 children (64%) with a significantly distended bladder had abnormal PVR. Ten of the 37 children (27%) without a significantly distended bladder had abnormal PVR (p ⫽ 0.02). Conclusions: Normal DVSS for gender is able to predict a bell-shaped uroflowmetry curve in most children, while EMG patterns and PVR do not correlate with uroflowmetry curves in these children. Children with a significantly distended bladder frequently have nonbell-shaped uroflowmetry curves as well as elevated PVR. KEY WORDS: bladder, urination disorders, electromyography, urodynamics

Dysfunctional voiding (DV) is a common clinical problem seen in approximately 40% of patients presenting to the pediatric urologist.1 Females are predominantly affected with a female-to-male ratio of 5:1. It is characterized by numerous symptoms, including recurrent urinary tract infections, urinary incontinence, constipation and encopresis. It is learned behavior that suppresses bladder contractions by inappropriately contracting the pelvic floor muscles (external urinary sphincter) during urination.2 This eventually becomes an involuntary process, resulting in functional obstruction of the urinary stream during voiding (detrusor/ external urinary sphincter dyscoordination). The Standardization Committee of the International Children’s Continence Society (ICCS) defined dysfunctional voiding as “over activity of the urethral sphincter during the voiding contraction of the detrusor in neurologically normal children.”3 Treatment options for this condition have included the selective use of anticholinergics, modification of elimination habits such as timed and multiple voiding, and biofeedback.4 The diagnosis of DV can be confirmed with limited urodynamics using uroflowmetry and electromyography.2, 3 The dysfunctional voiding symptom score (DVSS) was created as a noninvasive urodynamic alternative to reliably diagnose and monitor the treatment of children with DV.2 It

is a 10 question survey patterned after the International Prostate Symptom Score5 that translates quantitative and qualitative urological parameters into age appropriate questions for children. Receiver operating characteristics plots were used to define maximum normal threshold scores in 6 girls and 9 boys for optimal sensitivity and specificity. While the DVSS provides a numerical score that correlates with DV, to our knowledge it has never been directly correlated with detrusor/external urinary sphincter muscle activity. This current study was done to determine whether a normal DVSS can predict uroflowmetry and external urinary sphincter electromyography patterns in children. MATERIALS AND METHODS

A total of 55 children 4 to 13 years old were considered candidates for participation in our study. Approval from the university committee on research involving human subjects was obtained prior to patient recruitment. All candidates provided informed consent and completed a brief medical history emphasizing normal and abnormal voiding behaviors. In addition, a DVSS was completed by all patients with parental guidance. Individuals with known dysfunctional voiding, congenital urological abnormalities (except vesicoureteral reflux), neurological abnormalities or abnormal (above threshold) scores on the DVSS were excluded from study. Patients with scores normal for gender were considered eligible and scheduled for uroflowmetry/electromyography (EMG) and bladder scan for post-void residual urine volume (PVR) measurement.

Accepted for publication June 4, 2004. Study received university committee on research involving human subjects approval. Supported by a grant from the Meijer Corp., Grand Rapids, Michigan. 1980

DYSFUNCTIONAL VOIDING SYMPTOM SCORE AND UROFLOWMETRY

A detailed description of the uroflowmetry/EMG and bladder scan procedures was given to the patient and parent(s). Patients were given a choice of a male or female technician, each trained and experienced in uroflowmetry/ EMG, to perform the study. Parents were encouraged to stay with the child at the request of the child. Studies were performed in a separate and secure room in a pediatric urology office after regularly scheduled hours. Judicious use of gowns limited body exposure. The technician left the room during voiding at the request of the child to decrease potential anxiety. All children were encouraged to drink fluids and arrive for the study with a full bladder. The perineum and perianal skin were cleaned with a benzalkonium antiseptic towelette and dried with tissue paper. Pediatric adhesive electrode pads (1 inch in diameter) were placed immediately adjacent to the anus at the 3 and 9 o’clock positions. A UDS-600 Urodynamic System (Laborie Medical Technologies, Williston, Vermont) was used to collect data for uroflowmetry/EMG. The fluorescent lights were turned off to eliminate further any potential noise on the EMG recording. Baseline EMG was completed with the child in the right lateral recumbent position. After a baseline reading was obtained the child transferred to the uroflowmetry toilet seat and remained still without voiding until the EMG waveform stabilized. A footstool was provided to small children, so that the legs were supported. All children voided while seated. The child was then asked to relax and void without straining after the technician left the room. The electrodes were removed and the child was returned to the examination table, where PVR was measured while supine using an ultrasonic scanner (Diagnostic Ultrasound Corp., Redmond, Washington). Estimated bladder capacity (EBC) was determined using the formula, age in years ⫹ 2 ounces ⫻ 30 ml per ounce ⫽ EBC in ml.6, 7 A combined voided and PVR of 2/3 EBC was selected as a requirement for successful completion of the study. Children with a combined voided and PVR of less than 2/3 EBC were rescheduled for a second uroflowmetry/EMG attempt. All participants successfully completing the study were provided a gift certificate. Uroflowmetry waveforms were classified as bell-shaped, plateau (flattened) or fractionated (staccato) (fig. 1).3, 8 EMG activity was defined as decreased if there was sustained activity below resting baseline during voiding, variable if there was activity above and below resting baseline during voiding, and increased if there was sustained activity above resting base-

FIG. 1. Uroflowmetry waveforms

1981

FIG. 2. EMG tracings comparing resting baseline and voiding patterns.

line during voiding (fig. 2). Residual urine was considered abnormal if greater than 10% of EBC.2 For statistical analysis uroflowmetry curves were categorized as bell-shaped (normal) or nonbell-shaped (abnormal). Electromyography patterns were categorized as decreased (normal) or nondecreased (abnormal). Cohen’s ␬ was used to determine the level of agreement between uroflowmetry curves and EMG patterns, and uroflowmetry patterns and PVR.9 The ␬ score provides overall agreement between 2 measures accounting for the agreement that may occur by chance alone. A significantly distended bladder was defined as a combined voided plus PVR in excess of 30% above EBC. Fisher’s exact test was used to determine the significance of a distended bladder in children with bell-shaped vs nonbellshaped uroflowmetry curves and increased PVR in children with a significantly distended vs a nondistended bladder. RESULTS

A total of 55 children 4 to 13 years old were considered candidates for participation in the study. No children were eliminated due to preexisting urological problems. One girl was excluded due to an abnormal DVSS score of 7. Two children were excluded due to malfunction of the EMG unit and 1 was excluded due to loose stool during uroflowmetry. The remaining 51 participants comprised 17 boys and 34 girls with a mean age of 8.3 years (fig. 3). All DVSSs were below the minimum threshold for gender, as defined by Farhat et al.2 Scores ranged from 0 to 5 (mean 2.4 and 2.2 in boys and girls, respectively). A total of 37 children had bell-shaped uroflowmetry curves. In the 14 children with nonbell-shaped curves the curves were plateau-shaped in 8 and fractionated in 6 (fig. 4). Electromyography results consisted of increased patterns in 7 boys and 8 girls, variable patterns in 5 boys and 13 girls, and decreased patterns in 5 boys and 13 girls (fig. 5). Cohen’s ␬ was used to determine the strength of agreement between uroflowmetry curves and EMG patterns. A minimum of 0.70 indicates good agreement. The ␬ score for uroflowmetry curves and EMG patterns in our study was 0.11, showing poor agreement (fig. 6). Of 37 children with bell-shaped uroflowmetry curves 5 boys and 19 girls (65%) had normal PVR, while 6 boys and 7 girls (35%) had abnormal curves. Of the children with nonbell-shaped uroflowmetry curves 2 boys and 6 girls (57%) had normal PVR, and 3 boys and 3 girls (43%) had abnormal PVR (fig. 7). There was poor agreement between uroflowmetry curves and PVR (Cohen’s ␬ score 0.13).

1982

DYSFUNCTIONAL VOIDING SYMPTOM SCORE AND UROFLOWMETRY

FIG. 3. Age distribution of 51 participating children by gender

FIG. 4. Distribution of uroflowmetry curves by gender

A total of 14 children were considered to have a significantly distended bladder during uroflowmetry/EMG. This included 1 boy and 6 girls of the 37 children (19%) with bell-shaped curves and 3 boys and 4 girls of the 14 (50%) with nonbell-shaped curves (p ⫽ 0.04). Elevated PVR was seen in 5 boys and 5 girls of the 37 children (27%) with normal bladder volume and in 4 boys and 5 girls of the 14 children (64%) with a significantly distended bladder (p ⫽ 0.02). This suggests that elevated PVR is more common in children with a significantly distended bladder. DISCUSSION

Voiding abnormalities in children remain a common reason for referral to the pediatric urologist. Of these abnormalities are symptoms of urinary frequency and infrequency, urgency, incontinence, recurrent urinary tract infections and various bowel disturbances, including constipation and encopresis. While some children are diagnosed with a coexisting neurological abnormality, the majority are considered neurologically normal.

Descriptive terms such as nonneurogenic neurogenic bladder, occult neuropathic bladder, detrusor-sphincter dyssynergia and detrusor-sphincter dyscoordination have been used to explain the voiding abnormality. The variable terminology reflects the uncertain etiology of this condition and it has added to the confusion in the literature. However, a common finding is the functional obstruction of the urinary stream due to external urinary sphincter muscle contraction during urination. The ICCS has formally defined “over activity of the urethral sphincter during the voiding contraction of the detrusor in neurologically normal children” as dysfunctional voiding.3 The ICCS definition lends itself to the use of urodynamics for diagnosing children with DV. However, the use of invasive urodynamic techniques for evaluating voiding abnormalities in children has been abandoned in favor of noninvasive pediatric urodynamic techniques. While these methods have demonstrated reproducible results, a considerable amount of time, expense and manpower is required to obtain a quality study. Farhat et al devised the dysfunctional voiding scoring system questionnaire as a means of providing a quantitative

DYSFUNCTIONAL VOIDING SYMPTOM SCORE AND UROFLOWMETRY

1983

FIG. 5. Distribution of EMG patterns by gender

FIG. 6. Strength of agreement between uroflowmetry curves and EMG patterns

measurement for the diagnosis and monitoring of children with dysfunctional voiding.2 Receiver operating characteristics plots were used to define maximum normal threshold scores for 6 females and 9 males. While the DVSS provides a numerical score that correlates with the clinical definition of DV, to our knowledge it has never been directly correlated with detrusor/external urinary sphincter muscle activity. Because DV may be defined by urodynamics, we elected to compare children considered normal by DVSS using their uroflowmetry/EMG patterns. In general, uroflowmetry/EMG patterns in normal children have bell-shaped (parabolic) curves with decreased EMG activity during voiding.4, 8, 10 –13 In addition, post-void residual urine volumes are considered clinically significant when they are 20 or greater ml or greater than 10% of EBC.3 Our results show that 73% of children considered normal by DVSS criteria had bell-shaped uroflowmetry curves, while 27% had a flat or interrupted curve. In a study of urinary flow in healthy school children Mattsson and Spangberg determined that the normal urinary flow curve was bell-shaped regardless of sex, age or

voided volume.13 As a screening method, they believed that the shape of the flow curve was the most important factor to analyze. However, they also concluded that repeat deviations from the bell-shaped flow curve needed further investigation. Jensen et al found that more than 90% of children studied for the evaluation of uroflowmetry variables had flow curve patterns identical to the normal adult bell-shaped flow curve.11 However, they also identified that 68% of micturations diagnosed as dyssynergenic showed bell-shaped curves. Pelvic floor EMG has also been used to diagnose children with external urinary sphincter and dysfunctional voiding abnormalities. Our study results show that 65% of children who were normal by DVSS criteria had increased or variable EMG patterns during voiding. The ICCS describes the normal EMG pattern in children as a gradual increase in EMG activity from pelvic floor and urethral sphincter contraction during bladder filling with complete absence of activity at the onset of micturation.3 Any EMG activity from the sphincter or pelvic floor during voiding is abnormal unless the patient is attempting to inhibit micturation. Our study shows that

1984

DYSFUNCTIONAL VOIDING SYMPTOM SCORE AND UROFLOWMETRY

FIG. 7. Strength of agreement between PVR volumes and uroflowmetry curves

65% of normal children had increased or variable EMG patterns suggestive of DV. Koff and Kass considered that Valsalva type maneuvers and abdominal wall straining were a source of abnormal EMG artifact during voiding.14 They found that 59% of the children in their study strained to void, including 75% during the initiation of voiding and 64% during voiding. They also found that almost all children who strained could be coaxed into voiding without straining. While this straining may be a source of artifact, it may in actuality be part of the pathological complex contributing to dysfunctional voiding. Griffiths and Scholtmeijer suggested that, while EMG is intermediate in sensitivity, reproducibility and strength in association with abnormal voiding symptoms, it has an intermediate degree of over sensitivity to artifact.10 In our patient population of normal children the ICCS defined normal EMG pattern was identified in only 35% of patients, raising concern over its diagnostic ability. Repeating EMG tracings could help clarify this finding. Post-void residual urine measurement is a means of assessing lower urinary tract function and it was also a parameter studied in our children with a normal DVSS. Using the ICCS definition of abnormal PVR 35% of children with bellshaped uroflowmetry curves and 43% with nonbell-shaped curves had elevated PVR. Concern raised over the influence that a significantly distended bladder may have on PVR prompted a further breakdown of these children with elevated PVR. Of the 14 children with a significantly distended bladder 64% had an abnormal PVR, while 10 of 37 (27%) without a significantly distended bladder had an abnormal PVR. This suggests that increased PVR is more common in children with a significantly distended bladder. Overall 37% of the children had abnormal PVR. We believe that performance anxiety and the novel conditions in which the children were asked to void may have had an impact on all parameters measured. Concern over the influence that anxiety and straining may have on uroflowmetry/EMG and PVR results has also been raised by others.10, 11, 12, 14 Lidefelt et al found that even substantial residual urine volumes might occasionally occur in healthy children, suggesting a psychological influence on bladder control.15 Limiting the urodynamic studies to a single trial may also influence recorded outcomes. Our data raises the question of whether external urinary sphincter electromyography and PVR measurements in children represent a reli-

able screening test for dysfunctional voiding. No single test is likely to diagnose children with DV. A final explanation for the negative correlation between children with a normal DVSS and their uroflowmetry/EMG/ PVR results may be subclinical dysfunctional voiding. This might be a process whereby children with abnormal objective parameter measurements lack clinical symptoms. Plans for repeating the current study in 12 to 18 months using the same patient population could provide further insight into this. The DVSS is easy to use, inexpensive and noninvasive. In children with a normal score for gender the DVSS is able to predict a bell-shaped normal curve in 73% of cases. Excluding patients with a significantly distended bladder increased this to 81%. However, it is unable consistently to predict a typical EMG pattern or PVR. Further studies are needed to validate this questionnaire using normal and abnormal scores with uroflowmetry/EMG and PVR results. The DVSS promises to be a valuable tool for practicing urologists for diagnosing and managing DV in children. CONCLUSIONS

Bell-shaped uroflowmetry curves are the predominant flow curve in children with a normal DVSS score. EMG patterns and PVR do not correlate with uroflowmetry curves in children with normal DVSS scores. Children with a significantly distended bladder frequently have abnormal nonbell-shaped uroflowmetry curves as well as increased PVR. Nikki Bonter assisted with uroflowmetry and EMG, and Shannon Fries assisted with the manuscript. REFERENCES

1. Snodgrass, W.: Relationship of voiding dysfunction to urinary tract infection and vesicoureteral reflux in children. Urology, 38: 341, 1991 2. Farhat, W., Ba¨gli, D. J., Capolicchio, G., O’Reilly, S., Merguerian, P. A., Khoury, A. et al: The dysfunctional voiding scoring system: quantitative standardization of dysfunctional voiding symptoms in children. J Urol, 164: 1011, 2000 3. Norgaard, J. P., van Gool, J. D., Hjalamas, K., Djurhuus, J. C. and Hellstrom, A. L.: Standardization and definitions in lower urinary tract dysfunction in children. International Children’s Continence Society. Br J Urol, suppl., 81: 1, 1998 4. Porena, M., Costantini, E., Rociola, W. and Mearini, E.: Biofeed-

DYSFUNCTIONAL VOIDING SYMPTOM SCORE AND UROFLOWMETRY

5.

6.

7. 8. 9. 10. 11.

12. 13. 14. 15.

back successfully cures detrusor-sphincter dyssynergia in pediatric patients. J Urol, 163: 1927, 2000 Barry, M. J., Fowler, F. J., Jr., O’Leary, M. P., Bruskewitz, R. C., Holtgrewe, H. L., Mebust, W. K. et al: The American Urological Association symptom index for benign prostatic hyperplasia. J Urol, 148: 1549, 1992 Berger, R. M., Maizels, M., Moran, G. C., Conway, J. J. and Firlit, C. F.: Bladder capacity (ounces) equals age (years) plus 2 predicts normal bladder capacity and aids in diagnosis of abnormal voiding patterns. J Urol, 129: 347, 1983 Koff, S. A.: Estimating bladder capacity in children. Urology, 21: 248, 1983 Jorgensen, J. B. and Jensen, K. M.: Uroflowometry. Urol Clin North Am, 23: 237, 1996 Silcocks, P. B.: Measuring repeatability and validity of histological diagnosis—a brief review with some practical examples. J Clin Pathol, 36: 1269, 1983 Griffiths, D. J. and Scholtmeijer, R. J.: Detrusor/sphincter dyssynergia in neurologically normal children. Neurourol Urodyn, 2: 27, 1983 Jensen, K. M., Nielsen, K. K., Kristensen, E. S., Dalsgaard, J., Qvist, N. and Krarup, T.: Uroflowometry in neurologically normal children with voiding disorders. Scand J Urol Nephrol, 19: 81, 1985 Maizels, M., Kaplan, W. E., King, L. R. and Firlit, C. F.: The vesical sphincter electromyogram in children with normal and abnormal voiding patterns. J Urol, 129: 92, 1983 Mattsson, S. and Spangberg, A.: Urinary flow in healthy schoolchildren. Neurourol Urodyn, 13: 281, 1994 Koff, S. A. and Kass, E. J.: Abdominal wall electromyography: a noninvasive technique to improve pediatric urodynamic accuracy. J Urol, 127: 736, 1982 Lidefelt, K.-J., Erasmie, U. and Bollgren, I.: Residual urine in children with acute cystitis and in healthy children: assessment by sonography. J Urol, 141: 916, 1989 EDITORIAL COMMENTS

This provocative study should cast doubt on the use of noninvasive urodynamics for the evaluation and treatment of dysfunctional voiding in children. It also serves to emphasize the difficulty of applying mathematical parameters to the evaluation of bladder function. The authors studied a group of asymptomatic boys and girls without a history of urinary tract infection and found that a substantial number had abnormal test results. These children failed to void with smooth bell-shaped curves, failed to quiet EMG activity during voiding, had a distended bladder at capacity and had abnormally high post-void residual urine volumes. Therefore, the formula used to calculate normal bladder capacity for age is inaccurate. The judgment that some of these bladders were distended is also probably incorrect since it is likely that bladder capacities were not abnormal despite deviance from the formula. Perineal patch electrodes are notoriously fickle and they often show artifactual activity with movement or if moistened during voiding. Obviously many girls and boys can dysfunctionally void and somehow maintain continence and sterile urine. Therefore, it is probable that the etiology of incontinence and recurrent infection in children is multifactorial. It is also possible that noninvasive urodynamic studies fail to assess lower urinary tract function adequately. These studies do not measure voiding pressures, assess uninhibited bladder contractions, allow measurement of sphincteric pressures during voiding or enable one to visualize bladder wall or urethral morphology and look for vesicoureteral reflux. Despite its inherent shortcomings invasive

1985

fluoro-urodynamics may be necessary for the evaluation and treatment of some of these children. Saul P. Greenfield Department of Pediatric Urology Children’s Hospital of Buffalo Department of Urology State University of New York at Buffalo Buffalo, New York

In an attempt to streamline the ability to evaluate children with dysfunctional voiding in a noninvasive manner these authors applied a DVSS questionnaire developed at the pediatric urology service, Hospital for Sick Children, Toronto. The investigators wanted to learn if this questionnaire could be applied to the emptying phase of the micturition cycle in a group of children who were thought to be normal based on the DVSS score. Although 73% of children had a normal bell-shaped uroflowmetry curve and complete emptying, more than a quarter had an abnormal uroflowmetry curve, including a fractionated or staccato voiding pattern. The investigators found that only 35% of the children had total relaxation of EMG activity during voiding and only 65% emptied the bladder completely. Thus, there was little correlation between a normal DVSS score and how most of these children emptied the bladder. In fact, the ability to predict who would or would not empty the bladder was related only to the degree of bladder distention, as measured by voided volume and residual urine (p ⫽ 0.02). The authors used noninvasive methods of uroflowmetry, patch perineal electrodes and post-void residual urine measurement with an ultrasonic scanner to determine how normal children empty. Despite taking multiple precautions to ensure that the children were in as relaxed a state as possible (being seated, foot supports and privacy) clearly almost a third did not empty the bladder with what would be considered a normal uroflowmetry pattern. Even if one does not accept perineal patch electrodes as an accurate measure of urethral sphincter activity during voiding, the fact remains that incomplete emptying or staccato voiding was prevalent in a considerable number of children who were thought to be normal. As the authors clearly state, there are limitations in trying to identify children who are normal just through a questionnaire and I would share that sentiment. Urodynamic studies (even relatively noninvasive methods) are often needed to detect how children empty the bladder. Perhaps the DVSS questionnaire, which did not fully address voiding habits and uroflowmetry patterns as it was initially conceived, should be amended to incorporate the findings of these investigators better to improve its predictability for detecting normal and/or abnormal behaviors. This study confirms that the initial effort of the Toronto group needs further refinement. These authors are to be commended for meticulous attention to detail in trying to get the children as relaxed as possible when they voided for the test and in their analysis of the data acquired. As Bill Cosby related in a recent commencement day speech: 1 day after school when he asked his grandmother, “Is the glass half empty or half full?” she replied, “It depends on who’s pouring.”1 Stuart B. Bauer Department of Urology Children’s Hospital Harvard Medical School Boston, Massachusetts 1. Cosby, B.: Rice University graduation commencement speech, Houston, Texas, May 11, 2002