Factors Influencing Quality of Life in Children With Urinary Incontinence

Factors Influencing Quality of Life in Children With Urinary Incontinence Aniruddh V. Deshpande,* Jonathan C. Craig, Grahame H. H. Smith and Patrina H. Y. Caldwell From the Departments of Urology (AVD, GHHS) and Nephrology (JCC, PHYC), and Center for Kidney Research (AVD, JCC, PHYC), Children’s Hospital at Westmead, School of Public Health (AVD, JCC), Westmead, and Discipline of Pediatrics and Child Health, Sydney Medical School, University of Sydney, Sydney (PHYC), New South Wales, Australia

Abbreviations and Acronyms CSS ⫽ clinical severity score PinQ ⫽ Pediatric Incontinence Questionnaire QoL ⫽ quality of life Submitted for publication December 8, 2010. Study received human research ethics committee approval. * Correspondence: Center for Kidney Research, Children’s Hospital at Westmead, Locked Bag 4001, Corner Hainsworth St. and Hawkesbury Rd., Westmead, NSW, 2145, Australia (telephone: 61-2-9845-1463; FAX: 61-2-9845-3180, e-mail: [email protected]).

Purpose: We evaluated quality of life in children with urinary incontinence using a disease specific tool (Pediatric Incontinence Questionnaire) and determined factors that decrease quality of life in affected children. Materials and Methods: The Pediatric Incontinence Questionnaire was selfadministered by children 6 to16 years old with urinary incontinence while attending outpatient clinics at a tertiary pediatric hospital in Australia between October 2009 and May 2010. A weighted summative quality of life score with a range of 1.75 to 7 (7 being lowest quality of life) was generated, and patient characteristics (age, gender, ethnicity, symptom severity) were evaluated as potential predictors. Results: Of 146 children invited to participate 138 consented (response rate 95%). About half of the participants (77) were boys, and mean patient age was 10 years. Girls had a lower quality of life than boys (mean score 3.60 vs 3.31, 95% CI 0.10 – 0.57, p ⫽ 0.04), and nonwhite children had a lower quality of life than white children (3.97 vs 3.35, 95% CI 0.23– 0.99, p ⬍0.01). Older age (r ⫽ 0.21, p ⫽ 0.01) but not increasing symptom severity (r ⫽ 0.15, p ⫽ 0.09) or underlying chronic disease (difference 0.12, p ⫽ 0.91) was correlated to decreased quality of life. Multivariate regression analysis demonstrated that older age, nonwhite ethnicity and female gender were independent predictors of decreased quality of life. Conclusions: Older age, female gender and nonwhite ethnicity are associated with a lower disease specific quality of life in children with urinary incontinence. Clinicians need to be aware of the differential effect of urinary incontinence in children of different ages and ethnic backgrounds. Key Words: child, quality of life, risk factors, urinary incontinence

1048

www.jurology.com

URINARY incontinence is common in school-age children, with an estimated prevalence of 5% to 15%.1 The condition is a source of significant anxiety to children and their families. However, there are few data on the impact of urinary incontinence on quality of life in children.2– 6 Recently a disease specific quality of life assessment tool (Pediatric Incontinence Questionnaire) for bladder dysfunc-

tion in children has been developed and validated. The few published studies have highlighted a significant impact of urinary incontinence on quality of life in affected children.2,3 They have also indicated a differential effect of the nature of symptoms on various groups of children with urinary incontinence. Nevertheless, many aspects of the effect of urinary incontinence on quality of life in af-

0022-5347/11/1863-1048/0 THE JOURNAL OF UROLOGY® © 2011 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION

Vol. 186, 1048-1052, September 2011 Printed in U.S.A. DOI:10.1016/j.juro.2011.04.104

AND

RESEARCH, INC.

QUALITY OF LIFE IN CHILDREN WITH URINARY INCONTINENCE

fected children remain unreported. The effect of associated illnesses and underlying urogenital anomalies on perceived disease specific quality of life has been poorly studied. We describe QoL in children with urinary incontinence using PinQ, and we identify the predictors of disease specific decreased QoL in affected children. We also describe disease specific QoL in children with urinary incontinence in combination with urological and other chronic illnesses.

MATERIALS AND METHODS The study was approved by the human research ethics committee of the Children’s Hospital at Westmead. The hospital provides tertiary level care to children with urinary incontinence and bladder dysfunction through outpatient clinics run by a multidisciplinary team (incontinence physician, urologist, incontinence nurse, physiotherapist, psychologist).

Participants Children 6 to 16 years old with urinary incontinence who attended the clinic were invited to participate irrespective of type of urinary incontinence and response to treatment. Parental consent was obtained. Children with developmental delay and neurogenic bladder dysfunction were ineligible to participate in the study.

Survey Instrument and Administration Method PinQ is a disease specific instrument validated to measure the impact of bladder dysfunction on QoL in children with urinary incontinence.3 PinQ covers 7 domains using 28 questions that measure different psychometric attributes of importance to children with bladder problems. Each question has 4 possible responses that follow a Likert scale. PinQ was self-administered by the children. Parental assistance in interpreting the questions was permitted wherever necessary. Services of a qualified interpreter were available on request. In previous publications using this tool the total score has been calculated as a summation of all responses, with a higher score indicating a lower QoL.3,7 As a result of varying numbers of questions in each domain, domains were unequally represented in the final score. We chose to estimate the total score using equal weights for each domain since we could find no justification for superiority of one domain over another in the process of development and validation.2,3 Each domain was converted to a maximum score of 1, with a summed weighted maximum QoL score of 7 (minimum 1.75). A sensitivity analysis comparing the summation method to the weighted method was performed.

Clinical Data Collection and Severity Estimation In addition to QoL data, we collected data on patient gender, age, ethnicity and postal address (including post code), and details of disease and treatment from the routinely collected medical records. We estimated the concurrent severity of underlying bladder dysfunction using an instrument developed by Akbal et al.8 This instrument is calibrated between a minimum of 0 and a maximum of 35 points. A score of 8.5 or greater has been shown to be associated with voiding abnormalities. CSS was obtained

1049

from the summation of the responses to the 13 questions in the instrument. We used the Socioeconomic Indexes for Areas for relative socioeconomic advantage and disadvantage as prescribed by the Australian Bureau of Statistics to compare the socioeconomic status of the white and nonwhite groups (1 ⫽ lowest, 10 ⫽ highest, with 10 indicating better socioeconomic status).

Sample Size and Data Analysis Based on a previous publication,3 we estimated a total of 36 participants would be required in each subgroup based on symptomatology to detect a mean difference in a single domain score (range 1 to 16) of 1 standard deviation (SD ⫽ 3) with 80% power and a significance level of 0.05. A difference of 3 points in the summative score is approximately equal to a difference of 0.15 points in the weighted score used in this study. We anticipated 4 groups based on symptomatology, and hence we chose to include 144 participants in the study.

Model 1 All data were examined for spread and tested for normality. To identify independent risk factors associated with an adverse impact on QoL in children with urinary incontinence, we fitted a linear regression model that included individual variables and QoL score. Ethnicity and gender were treated as categorical data. Due to the small number of children in each nonwhite ethnic group, these cases were collapsed into a single category to facilitate analysis. Age, CSS and QoL score were treated as continuous data. Effect modification due to gender and ethnicity was excluded. Backward selection technique was used to arrive at a final model. A p value of 0.05 or less was considered significant.

Model 2 We further tested whether the presence of other underlying chronic conditions (urological or other chronic diseases) reduced the impact of urinary incontinence on QoL in children after adjusting for other independent predictors of QoL. Participants were collapsed into 2 categories based on the presence of underlying disease. Standard multiple regression was used to check whether these categories revealed a statistically significant association with QoL scores after adjusting for other important variables as identified in model 1. Both models were tested for validity by examining the residuals. Participants were excluded if 7 responses (25%) or more were missing. In those with fewer than 7 missing responses missing values were replaced by series means. Data were analyzed using SAS®, version 9.1.

RESULTS Of 146 children invited to participate in the study 144 consented. Six questionnaires were excluded from analysis (2 that were blank, 2 with no identifiers and 2 with more than 20 missing values). Characteristics of the 138 remaining participants are outlined in table 1. The majority of patients were white. Nearly a third of children participated in the study at the first visit to the clinic. Most nonwhite participants were of Middle Eastern origin. Approximately a fourth of participants had an underlying

1050


chronic illness in addition to urinary incontinence. Mean ⫾ SD decile value of the Socioeconomic Indexes for Areas score for assessment of socioeconomic status was 5.78 ⫾ 2.6 (median 5.5) for nonwhites and 6.1 ⫾ 2.9 (6.0) for whites. Underlying chronic illnesses were attention deficit/ hyperactivity disorder (7 children), cystic fibrosis (2), congenital heart disease (2), type 1 diabetes mellitus (2), obesity (3) and asthma (4). Urological conditions included vesicoureteral reflux (5 children), posterior urethral valves (3), complex genital anomalies including bladder exstrophy (2), posttraumatic incontinence (1) and bilateral ectopic ureters (1). Mean ⫾ SD CSS for the study group was 9.75 ⫾ 5.7. Quality of Life The sensitivity analysis comparing the composite summative score as reported previously with the weighted estimate suggested that the 2 scores were proportionately similar, and hence only data from the equally weighted domain method are reported. Mean ⫾ SD QoL score for the cohort was 3.44 ⫾ 0.82 (range 1.75 to 5.66, table 2). Mean domain specific, pretreatment scores for the 40 children whose participation in the study coincided with their first visit to the clinic are listed in table 3. Self-esteem and mental health were the most affected domains (p ⬍0.01, one-way ANOVA). Risk Factors for Decreased Quality of Life On univariate analysis females had a higher mean QoL score than males (t134 ⫽ ⫺2.07, table 4). Similarly children of nonwhite ethnicity had higher scores than white children (t136 ⫽ ⫺3.23). QoL scores increased with increasing age (r ⫽ 0.21) but not with increasing CSS (r ⫽ 0.15). Multivariate analysis in the base model suggested that increasing age and ethnicity were significantly associated with QoL score, while CSS Table 1. Patient characteristics No. Pts (%) Age (yrs): Younger than 8 8–11 12 Or older Gender: M F Ethnicity: White Middle Eastern Chinese Indian/Sri Lankan Condition: Monosymptomatic nocturnal enuresis* Daytime wetting (with or without enuresis)† * Mean ⫾ SD CSS 7.6 ⫾ 4.5 (range 0 to 35). † Mean ⫾ SD CSS 11.9 ⫾ 5.9 (range 0 to 35).

48 (34.8) 49 (35.5) 41 (29.7) 75 (54.4) 63 (45.6) 118 (85.5) 14 (10.14) 3 (2.18) 3 (2.18) 45 (32.6) 39 (28.3)

Table 2. QoL scores Mean ⫾ SD Score* Gender: M F Ethnicity: White Nonwhite Condition: Monosymptomatic nocturnal enuresis Daytime wetting (with or without enuresis) Urological illnesses with urinary incontinence Chronic illnesses with urinary incontinence

3.31 ⫾ 0.75 3.60 ⫾ 0.88 3.35 ⫾ 0.79 3.97 ⫾ 0.81 3.34 ⫾ 0.83 3.38 ⫾ 0.77 3.23 ⫾ 0.59 3.68 ⫾ 0.90

* Weighted summative domain scores using PinQ instrument. Each domain is weighted as maximum score of 1. Scores ranged from 1.75 to 7.

(p ⫽ 0.07) and gender (p ⫽ 0.06) were not. There was no effect modification. In the final model gender, age and ethnicity were significantly associated with weighted QoL score, and thus were independent predictors of QoL in these children (table 4). The final model accounted for about 14.4% of the variability in QoL score. Presence of underlying illnesses was not significantly associated with QoL scores on univariate or multivariate analysis. Mean QoL scores for the categories bladder dysfunction only (103 patients, mean ⫾ SD score 3.43 ⫾ 0.82) and bladder dysfunction with underlying illnesses (35, 3.46 ⫾ 0.86) were similar (p ⫽ 0.91). When added to the multiple regression model mentioned, the difference continued to be statistically nonsignificant (p ⫽ 0.80).

DISCUSSION Due to the growing recognition of patient perspectives in health care, quality of life assessment is an important part of incontinence research.5,7 Measurement of quality of life in children with urinary incontinence gives a child centric estimate of the impact of urinary incontinence. Quality of life can be measured using generic or disease specific tools. We chose a disease specific, cross-cultural QoL metric because such tools appear Table 3. Individual PinQ domains at first visit Domain

Mean ⫾ SD Score

Social relations with peers Self-esteem Family/home Body image Independence Mental health Treatment

0.436 ⫾ 0.16 0.610 ⫾ 0.22 0.518 ⫾ 0.15 0.484 ⫾ 0.18 0.467 ⫾ 0.15 0.557 ⫾ 0.22 0.479 ⫾ 0.16


1051

Table 4. Patient characteristics and overall QoL scores Univariate Analysis Est QoL Score (95% CI) Age (increase/yr) Gender: M F Ethnicity: Nonwhite White CSS (increase/unit)

0.068

(0.014–0.122)

0.279

Referent (0.007–0.552)

Multivariate Analysis p Value 0.01 0.04

Est QoL Score (95% CI) 0.068

(0.016–0.119)

0.293

Referent (0.035–0.551)

⬍0.01 Referent ⫺0.617 (⫺0.995–⫺0.240) 0.021 (-0.003–0.046)

to have certain advantages over generic QoL instruments,9 given their better content validity and psychometric properties when used for adults with urinary incontinence. In our study self-esteem and mental health were the 2 most affected domains in children with urinary incontinence, in agreement with previous series.3,5,10,11 Improvement in selfesteem has been reported after successful treatment in children with nocturnal enuresis as well as daytime wetting.10,11 The severity of the underlying bladder problem is expected to have a proportional adverse effect on QoL in children with urinary incontinence. Hence, incorporation of a concurrent quantitative measure of symptom severity into the regression model is a major strength of our study design. We chose a scale with wide applicability,8 instead of disease specific scales.12,13 The estimated clinical severity was not statistically significantly correlated to QoL score in either univariate or multivariate analysis. Although not often recognized by parents or clinicians, this finding serves to highlight that even mild urinary incontinence can have a significant impact on QoL in affected children. It is also possible that the PinQ instrument has decreased sensitivity in children with a combination of bladder problems and other chronic illnesses, resulting in spuriously lower scores (better QoL) in these children. Our study identifies the characteristics of children most likely to report a severe psychosocial impact of their bladder problem. Nonwhite children have a lower QoL than their white counterparts with a similar severity of disease. In a cross-cultural study Bower found no overall difference between cultural groups.3 Nevertheless, she also observed a higher self-esteem score (and worse QoL) in European vs Asian children. It is noteworthy that the white and nonwhite participants in our study had similar socioeconomic scores. Hence, it is possible that the lower QoL in nonwhites reflects different cultural values or unrecognized difficulties in understanding the English questionnaire among children of Middle Eastern origin, although no parents requested the help of a

0.09

p Value 0.01 0.03

⬍0.01 Referent ⫺0.607 (⫺0.972–⫺0.242) —

—

qualified interpreter. However, selection bias due to a small nonwhite sample cannot be excluded. It is also possible that this difference is due to the modes of administration of the questionnaire, which was administered by a clinician in the referent studies,3,7 compared to our self-administered questionnaire. Female gender was associated with a worse QoL in univariate and multivariate analysis. To our knowledge these findings have not been reported before. Other investigators have reported no significant differences in the scores between boys and girls,3,7 which differs from the finding that boys had a significantly worse score for the self-esteem domain in 1 study when the questionnaire was administered orally by the physician. We found that the impact of urinary incontinence tends to be higher in older children. Surprisingly the earlier studies that used the same tool did not demonstrate this association.3,7 This finding is important given the common belief that urinary incontinence tends to resolve with age. It is possible that older children need more psychological support despite relatively minor symptoms. These findings have important treatment implications for identifying and targeting children at high risk for being severely affected by their incontinence. Urinary incontinence is being increasingly reported in a variety of common conditions of childhood, such as cystic fibrosis, asthma, obesity, diabetes, urological diseases and other chronic conditions.6,14 –18 These diseases together account for a large proportion of chronic diseases of childhood in the community. We have the advantage of including such cases since our referral sources include pediatric subspecialists as well as general pediatricians. Our results suggest that there is no difference in QoL between children with and without chronic conditions even after adjusting for other variables. This finding, albeit surprising, is consistent with a study by Dodson et al, which showed that the impact of urinary incontinence in children with chronic kidney disease is high and independent of the severity of underlying kidney disease.6 The results of our study provide useful leads for future research. The difference in perceived QoL be-

1052


tween boys and girls needs to be explored further. The validity of PinQ in children with associated illnesses needs to be studied better to enhance its applicability.

CONCLUSIONS The results of this study are a valuable addition to the existing data on the impact of urinary incontinence on QoL in children. It appears that QoL is worse in girls and children of nonwhite ethnicity, and worsens with age. Incontinence severity does not increase the impact or presence of underlying diseases, or reduce the impact of urinary inconti-

nence on QoL in affected children. In addition, our findings will serve to increase the awareness of underlying urinary incontinence by health care staff dealing with other chronic conditions such as obesity, diabetes, cystic fibrosis and urological anomalies. We hope that these findings will help set better treatment goals and be useful in identifying individual factors that can impair QoL. We also hope that they will enable treating clinicians to arrange simultaneously for counseling or support by qualified individuals to minimize the adverse impact on the well-being of these children and their families.

REFERENCES 1. Sureshkumar P, Craig JC, Roy LP et al: Daytime urinary incontinence in primary school children: a population-based survey. J Pediatr 2000; 137: 814. 2. Bower WF, Wong EM and Yeung CK: Development of a validated quality of life tool specific to children with bladder dysfunction. Neurourol Urodyn 2006; 25: 221. 3. Bower WF: Self-reported effect of childhood incontinence on quality of life. J Wound Ostomy Continence Nurs 2008; 35: 617. 4. Bachmann C, Lehr D, Janhsen E et al: Health related quality of life of a tertiary referral center population with urinary incontinence using the DCGM-10 questionnaire. J Urol, suppl., 2009; 182: 2000. 5. Gladh G, Eldh M and Mattsson S: Quality of life in neurologically healthy children with urinary incontinence. Acta Paediatr 2006; 95: 1648. 6. Dodson JL, Cohn SE, Cox C et al: Urinary incontinence in the CKiD cohort and health related quality of life. J Urol, suppl., 2009; 182: 2007. 7. Bachmann C, Lehr D, Janhsen E et al: German version of the Pediatric Incontinence Question-

naire for urinary incontinence health related quality of life. J Urol, suppl., 2009; 182: 1993. 8. Akbal C, Genc Y, Burgu B et al: Dysfunctional voiding and incontinence scoring system: quantitative evaluation of incontinence symptoms in pediatric population. J Urol 2005; 173: 969. 9. Oh SJ and Ku JH: Does condition-specific quality of life correlate with generic health-related quality of life and objective incontinence severity in women with stress urinary incontinence? Neurourol Urodyn 2006; 25: 324. 10. Hagglof B, Andren O, Bergstrom E et al: Selfesteem before and after treatment in children with nocturnal enuresis and urinary incontinence. Scand J Urol Nephrol Suppl 1997; 183: 79.

13. Sureshkumar P, Cumming RG and Craig JC: Validity and reliability of parental report of frequency, severity and risk factors of urinary tract infection and urinary incontinence in children. J Urol 2006; 175: 2254. 14. Nixon GM, Glazner JA, Martin JM et al: Urinary incontinence in female adolescents with cystic fibrosis. Pediatrics 2002; 110: e22. 15. Prasad SA, Balfour-Lynn IM, Carr SB et al: A comparison of the prevalence of urinary incontinence in girls with cystic fibrosis, asthma, and healthy controls. Pediatr Pulmonol 2006; 41: 1065. 16. Schwartz B, Wyman JF, Thomas W et al: Urinary incontinence in obese adolescent girls. J Pediatr Urol 2009; 5: 445.

11. Natale N, Kuhn S, Siemer S et al: Quality of life and self-esteem for children with urinary urge incontinence and voiding postponement. J Urol 2009; 182: 692.

17. Vella M, Cartwright R, Cardozo L et al: Prevalence of incontinence and incontinence-specific quality of life impairment in women with cystic fibrosis. Neurourol Urodyn 2009; 28: 986.

12. Farhat W, Bägli DJ, Capolicchio G et al: The dysfunctional voiding scoring system: quantitative standardization of dysfunctional voiding symptoms in children. J Urol 2000; 164: 1011.

18. Sarma AV, Kanaya AM, Nyberg LM et al: Urinary incontinence among women with type 1 diabetes— how common is it? J Urol 2009; 181: 1224.

EDITORIAL COMMENT This is a good study that addresses an important health outcome. The authors reveal that there are some factors that may decrease or impact QoL in children who suffer from urinary incontinence, including increasing age, female gender and nonwhite ethnicity. Nevertheless, it is becoming evident that QoL is multifactorial and may be dependent not only on the disease per se, but also on socioeconomic, cultural and possibly patient cognitive power. It is also apparent that it is important to seek the perspective of the child directly to determine QoL adequately. Understandably these multiple factors are not easily reported and may need to be addressed in future studies.

The fact that older patients are able to report lower quality of life is somehow intuitive since, understandably, as patients age, incontinence somehow becomes more of a psychological issue. Furthermore, as the age of the child increases, we naturally would expect patients to be able to express themselves better, which may explain why we see the trend that increasing age impacts quality of life. Walid A. Farhat Pediatric Urology Hospital for Sick Children (SickKids) University of Toronto Toronto, Ontario, Canada