SEARCH

SEARCH BY CITATION

Keywords:

  • psychometrics, quality of life;
  • translations;
  • urinary incontinence

ABSTRACT

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. ACKNOWLEDGMENTS
  9. APPENDIX
  10. REFERENCES

Aims

The Urogenital Distress Inventory (UDI-6) and Incontinence Impact Questionnaire (IIQ-7) assess symptom distress and the impact on daily life of urinary incontinence. The UDI-6 has not been validated before in males. Our aim was to validate the UDI-6 and IIQ-7 in Dutch men and women.

Methods

The translation to Dutch followed standardized procedures. We validated the IIQ-7 with and without an additional gender-neutral item (IIQ-SF). Adults with urinary incontinence for at least 3 months, completed the measures at inclusion; 1-week after inclusion to evaluate the test-retest reproducibility; and 6 months after inclusion with the addition of the RAND-36 health transition item to assess responsiveness and interpretability. To assess the discriminate ability, a reference population was enrolled. To assess construct validity, the urodynamic diagnosis was used.

Results

Questionnaire data of 160 patients were analyzed. Patients reported more symptoms and bother than the reference population (P < 0.001). The internal consistency was good in the IIQ-SF baseline scores (Cronbach's alphas 0.86–0.92), though moderate in the UDI-6 (Cronbach's alphas 0.44–0.66). Both measures showed good reproducibility at the test-retest (Intraclass Correlations Coefficients 0.75–0.85). Construct was adequate with 75% confirmed hypotheses of urodynamic data with measure scores. The measures were responsive after treatment with smaller measurement errors than the minimal important change. No floor or ceiling effects were observed in baseline data.

Conclusions

The Dutch UDI-6 and IIQ-7 are reliable, valid, and responsive instruments for assessing symptom distress of urinary incontinence and its impact on daily life in both men and women. Neurourol. Urodynam. 34:24–31, 2015. © 2013 Wiley Periodicals, Inc.


Abbreviations
AUC

area under the ROC curve

DO

detrusor overactivity

GLM

general linear model

HRQoL

health related quality of life

ICC

intraclass correlation coefficient

IIQ

Incontinence Impact Questionnaire

IIQ-7

seven-item Incontinence Impact Questionnaire

IIQ-8

eight-item Incontinence Impact Questionnaire

IIQ-SF

incontinence Impact Questionnaire Short Forms: IIQ-7 original, IIQ-7 adjusted and IIQ-8

LOA

limits of agreement

LUTS

lower urinary tract symptoms

MIC

minimal important change

RAND-36-HTI

health transition item of the RAND 36-item health survey

ROC

receiver operating characteristic

SD

standard deviation

SUI

stress urinary incontinence

UDI

urogenital distress inventory

UDI-6

six-item urogenital distress inventory

UI

urinary incontinence

INTRODUCTION

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. ACKNOWLEDGMENTS
  9. APPENDIX
  10. REFERENCES

Urinary incontinence (UI) affects approximately 30% to 60% of women and 1% to 39% of men, depending on age, aetiology, and the definition of incontinence used.[1] Various measures are available to assess the impact of UI on health related quality of life (HRQoL). Generic HRQoL measures lack sensitivity to the unique aspects of a specific disease,[2, 3] therefore disease-specific HRQoL measures can be considered more applicable in capturing the impact of a specific disease such as UI. These measures may also serve to assess the impact of treatment on UI and to facilitate future research in incontinence treatment.

The urogenital distress inventory (UDI) and incontinence impact questionnaire (IIQ) were both developed to assess the impact of UI on HRQoL.[4, 5] Short versions of the UDI and IIQ, UDI-6, and IIQ-7, were developed to reduce the respondents' burden.[6] The UDI-6 and IIQ-7 are both “A grade” recommended by the Fourth International Consultation on Incontinence.[7] A measure that is valid and reliable for a particular language and culture may not prove so when used in a different population.[8] The UDI and IIQ long forms were previously translated and revisited in Dutch,[9] and are nowadays widely used in the Netherlands in urogynecology clinical practice and research.[10-12] An unvalidated Dutch translation of the UDI-6 and IIQ-7 is also commonly used.[13, 14] However, by using a questionnaire without the measurement properties tested,[15] it remains unknown whether a reliable, valid, and responsive measure for the population of interest is used. Though not yet in Dutch, the UDI-6 and IIQ-7 have been translated and validated in other languages.[16-22] Although both measures[6] were developed for female patients, we aimed to validate these in both sexes. Unlike the UDI-6, the IIQ-7 has earlier been validated in men.[23] The objective of this study was to validate the UDI-6 and IIQ-7 in Dutch to provide a useful evaluation tool for the use in both men and women suffering from UI.

MATERIALS AND METHODS

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. ACKNOWLEDGMENTS
  9. APPENDIX
  10. REFERENCES

This observational study was conducted at a tertiary Urology and Gynaecology centre as part of a larger validation study of HRQoL pelvic floor measures. The study was approved by the Institutional Ethical Committee (MEC-2008-376).

Questionnaires

The UDI-6 is a six item symptom inventory, specific to symptoms associated with lower urinary tract dysfunction, and combines information on irritative, stress and obstructive/discomfort symptoms.[6] The IIQ-7 is a seven item life-impact assessment instrument specific to UI, and covers separate domains of physical activity, travel, social/relationships, and emotional health.[6] Both measures were developed for self-administration and are intended to be used in combination. Patients rated how much they experience impaired function of UI and the extent to which this UI affects their daily functioning with four response options per item ((0) “not at all”; (1) “slightly”; (2) “moderately”; (3) “greatly”). We maintained the scoring procedure as described by the original developer: if more than two items are missing, a total score is not to be calculated.[6] The mean score of items is multiplied by 33 1/3 to convert to a 0–100 scale.[6] Higher scores indicate more symptom distress (UDI-6), or more impact on daily life (IIQ-7).

In consultation with the original developer we concluded that the UDI-6 had good face validity concerning male patients, that is, the questions seemed plausible, relevant, and to span the domain adequately. In the IIQ-7 however, the item addressing household chores (“Has urine leakage affected your ability to do household chores (cooking, housecleaning, laundry)?”) might be related more to females. Therefore, based on the factor loadings of the IIQ long form[4] and as suggested by the original developer, we added a question about employment (“Has urine leakage affected your employment (work) outside the home?”). The addition of this “employment” item resulted in three versions of the IIQ-7 completed by both men and women: (1) IIQ-7 original [6]; (2) IIQ-7 adjusted; in which the item addressing “household chores” was substituted by the item “employment” (“Has urine leakage affected your employment (work) outside the home?”)[4] (3) IIQ-8; in which the item “employment” was added to the original IIQ-7. We refer to these versions as the IIQ-short forms (IIQ-SF). Referring to the IIQ long form[4], the “employment” item was part of the “travel” domain while the “household chose” item was part of the “physical activity” domain. However, unlike the IIQ long form, the IIQ-7 does not contain separate domain scores,[6] and therefore only total scores of the IIQ-SF are reported.

Linguistic Validation

The translation of the measures to Dutch followed standardized forward-backward procedures[8]: three independent forward translations, and backward translation by a native speaker. The Dutch versions of the measures were tested in 10 patients with UI, where potential problems were explored and discussed guided by a checklist. UI was defined according to the International Continence Society, thus “the complaint of any involuntary leakage of urine”.[24] As a result, the lay-out was adapted to clarify the instructions on how to indicate the chosen answer. Several minor problems were identified which did not indicate a need to adapt the content. The Dutch versions of UDI-6 and IIQ-SF were then finalized (Appendix).

Study Population and Study Design

Inclusion criteria included males and females aged 18 years or older, symptoms of UI for at least 3 months, and fluent and literate in the Dutch language. Exclusion criteria were symptomatic urinary tract infections, neurologic diseases (except diabetic neuropathy), active malignant tumours, dementia, and mental retardation. At the initial physician office visit, the treating physician explained the study to all consecutive patients potentially eligible for inclusion. The physician logged the name and gender of patients who were interested in participation, and gave an information package including two sets of questionnaires. These patients were phoned by the principal investigator for further explanation, and were instructed to complete the first set of questionnaires (baseline) at home, and to return it by postal mail directly after completion, together with a signed informed consent form. The second set of questionnaires were completed 1-week after baseline, and returned by postal mail. Six months after baseline, patients received through postal mail a third set of questionnaires to be completed at home and returned by postal mail. Age and education were documented through the questionnaire. Educational level was classified as “low” (primary school), “intermediate” (high school) or “high” (college or university degree). The test-retest period of 1 week between the repeated administrations was long enough to prevent recall, though short enough to ensure that clinical change had not occurred. During this period no treatment was initiated or changed. Given the observational nature of the study, the treating physician was free to perform any investigation (e.g., urodynamic studies) and prescribe any treatment after completion of the second round of questionnaires, for individual patient care.

The third questionnaire included the health transition item of the RAND 36-Item Health Survey (RAND-36-HTI),[25] in which patients were asked to score the change in their general health compared to 1-year ago. The response options were: (1) “much better;”(2) “a little better;” (3) “same;” (4) “a little worse;” (5) “much worse.” This RAND-36-HTI was used as the anchor (external criterion)[26] for the evaluation of responsiveness and interpretability.

Urodynamic Studies

Because of the observational nature of this study, multichannel urodynamic testing was only carried out if indicated according to the standard evaluation protocol of our clinic as recommended by the Fourth International Consultation of Incontinence Recommendations.[27] Urodynamic testing was performed according to the International Continence Society standards.[28] Using urodynamic studies, patients were categorized into groups: no urodynamic abnormality detected, stress urinary incontinence (SUI), detrusor overactivity (DO), and detrusor underactivity. When mixed SUI/DO incontinence was detected the type of treatment patient received during follow-up determined the classification for analysis.

Reference Group

Baseline data from a reference group were collected using a random subsample of an ISO-certified (ISO 26362) Dutch panel.[29] This subsample was stratified by gender, age, educational level and residential area, and therefore representative for the Dutch population above the age of 18. No medical data was available for this internet panel. A total of 450 panel participants were invited in order to meet our targeted sample size of 250 participants.

Statistical Methods

Statistical analysis was conducted using IBM® SPSS-software 20.0. Statistical significance was defined as P-value <0.05. The mean and standard deviation (SD) are reported for continuous data. For discrete data, the count and percentage are reported.

As described by the original developer,[6] if one or two items were missing in the measures, they were replaced by the average of the respondent's observed values (unconditional mean imputation[30]). The UDI-6 or IIQ-7 scores were not calculated if more than two items were missing.[6]

To assess the differences between patient and reference group, the Student's t-test was used for continuous data, and the Chi-squared test for categorical variables. General linear models were used to compare measure scores, controlling for variables that differed significantly between the patient and reference group in univariate analysis.

The Welch's F statistic was used as a single global test to compare means between groups, since homogeneity was not assumed. In case of significant differences between means, pairwise comparisons using Games–Howell analysis was performed.

For receiver operating characteristics (ROC) analysis the anchor RAND-36-HTI was dichotomized: patients who reported that they were “a little better” and “much better” were classified as “improved,” while “same,” “a little worse,” and “much worse” were classified as “not improved.”

We tested the measurement properties of the measures[31]:

  • The internal consistency, that is, the extent to which the items are measuring the same underlying construct, was tested by calculating Cronbach's alpha using baseline scores. A Cronbach's alpha between 0.70 and 0.95 was considered to reflect good internal consistency.[31]
  • The reproducibility, that is, the degree to which repeated measurements in the test-retest period provide similar answers, concerns reliability, and agreement. The reliability, that is, the extent to which patients can be distinguished from each other despite measurement errors, was calculated with the Intraclass Correlation Coefficient (ICC) according to McGraw and Wong for agreement,[32] and was considered acceptable when the ICC was ≥0.70.[31] Agreement, that is, the extent to which the scores on repeated measures are close to each other, was quantified using the limits of agreement (LOA) as described by Bland and Altman.[33] The LOA were calculated as the absolute mean change in scores (meanchange) of repeated measurements during the test-retest period ± 1.96 × SD (SDchange). The LOA estimates where 95% of individual differences fall, that is the absolute measurement error. Agreement is considered good if the LOA are smaller than the minimal important change (MIC; see interpretability).[31]
  • The construct validity, that is, the extent to which the scores relate to other measures, was verified by urodynamic diagnosis for discriminative validation. We formulated specific hypotheses and expected at least 75% of the results to be in accordance with these hypotheses[31]:
    • We expect patients without an abnormality detected on urodynamic studies, to have significantly better IIQ-SF scores than patients with a urodynamic diagnosis;
    • We expect patients with DO to score significantly worse in the irritative domain of the UDI-6, than patients with other urodynamic diagnoses;
    • We expect patients with SUI to score significantly worse in the stress domain of the UDI-6, than patients with other urodynamic diagnoses;
    • We expect patients with dysfunctional voiding or underactive detrusor, to score significantly worse in the obstructive domain of the UDI-6, than patients with other urodynamic diagnoses.
  • To assess responsiveness, that is, the ability of a measure to detect clinically important changes over time, in treated patients we firstly evaluated the linear relationship of the meanchange in measure scores between baseline and 6-month follow-up with the RAND-36-HTI score, using the Pearson correlation coefficient. In addition, the area under the ROC curve (AUC) for the UDI-6 and IIQ-SF measures was determined. The AUC indicates the probability that a measure correctly classified patients as improved, using the RAND-36-HTI as an anchor. The AUC was considered adequate if ≥0.70.[15]
  • To assess interpretability, that is, the degree to which one can assign qualitative meaning to quantitative scores, the anchor based ROC method was used to assess the MIC.[15] The MIC was defined as the optimal ROC cut-off point, that is, the value for which the sum of the proportions of misclassifications ([1 − sensitivity] + [1 − specificity]) is smallest.[15]
  • Floor and ceiling effects occur when high proportions of respondents report scores at the lower or upper end of the scale, indicating limited content validity, and thus may lead to a reduced reliability and limited responsiveness.[15] Floor or ceiling effects are considered to be present if more than 15% of respondents achieve the lowest or highest possible score.[31]

RESULTS

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. ACKNOWLEDGMENTS
  9. APPENDIX
  10. REFERENCES

Of 240 male and female consecutive patients, 218 were found to be eligible for inclusion (Fig. 1). Of the 218 patients, 174 (80%) consented to participate. After three rounds of questionnaires data were available for 160 out of 174 patients (92%) to test at least one measurement property (i.e., patients completed the questionnaires at baseline and at least at one additional time-point) of at least one questionnaire (i.e., patients completed the UDI-6 and/or IIQ-7). Regarding the reference group, the measures were sent out to 450 panel participants of which 279 (62%) responded.

image

Figure 1. Study flow-chart.

Download figure to PowerPoint

The study population had a mean age of 62 ± 12 years, and 59% were female (Table I). At baseline, patients reported significantly more symptoms of UI (UDI-6, P < 0.001) and more impact of these symptoms on daily life (IIQ-7 original, IIQ-7 adjusted, and IIQ-8 all P < 0.001) than the reference group. After adjusting for age and educational level, these differences remained significant (P < 0.001).

Table I. Characteristics of the Respondents and Questionnaires Scores
 Patients (n = 160)Reference group (n = 279)P-value univariateP-value GLM*
  • Data in mean ± standard deviation or number (%).

  • a

    Missing n = 2.

  • b

    Higher scores indicate more symptom distress (UDI-6) or more impact on daily life (IIQ-SF).

  • *

    P-value: corrected for age and educational level with general linear model (GLM).

Age (years)62 ± 1249 ± 16<0.001 
Gender  0.08 
Males65 (41%)138 (50%)  
Females95 (59%)141 (51%)  
Educational levela  <0.001 
Low59 (37%)21 (8%)  
Intermediate73 (46%)172 (62%)  
High28 (18%)86 (31%)  
Scores at baseline (0–100)b    
UDI-646.4 ± 16.812.2 ± 12.7<0.001<0.001
IIQ-7 original38.4 ± 24.64.0 ± 10.9<0.001<0.001
IIQ-7 adjusted39.4 ± 25.44.2 ± 11.2<0.001<0.001
IIQ-837.3 ± 24.212.2 ± 12.7<0.001<0.001

Internal Consistency

In the patient group, Cronbach's alpha was 0.49 for the UDI-6 total score and the domains ranged between 0.46 and 0.65, indicating moderate internal consistency (Table II). In the reference group, Cronbach's alpha was 0.66 for the UDI-6 total score, and the domains ranged between 0.44 and 0.58, also indicating moderate internal consistency. Internal consistency remained moderate after stratification by gender (not in Table).

Table II. Internal Consistency and Reproducibility
 PatientsReference group (n = 279)Patients test-retest reliability
Internal consistency (Cronbach's alpha)ICCagreementMeanchange ± SDchangeLimits of agreementa
  • All scale scores are transformed to a 0–100 point scale. Higher scores indicate more symptom distress (UDI-6) or more impact on daily life (IIQ-SF).

  • a

    Limits of agreement described by Bland and Altman[33] = meanchange ± 1.96 × SDchange.

UDI-60.49 (n = 149)0.660.84 (n = 152)6.3 ± 7.0−7.42 to 20.0
Irritative0.65 (n = 155)0.530.83 (n = 152)8.9 ± 13.1−16.8 to 34.6
Stress0.46 (n = 152)0.580.83 (n = 152)8.4 ± 11.5−14.1 to 30.9
Obstructive/discomfort0.54 (n = 154)0.440.85 (n = 152)7.3 ± 11.8−15.8 to 30.4
IIQ-7 original0.87 (n = 147)0.910.76 (n = 153)9.9 ± 14.1−17.7 to 37.5
IIQ-7 adjusted0.86 (n = 132)0.910.77 (n = 150)10.2 ± 14.1−17.4 to 37.8
IIQ-80.87 (n = 131)0.920.75 (n = 149)10.0 ± 14.1−17.6 to 37.6

The Cronbach's alpha of the IIQ-7 original, IIQ-7 adjusted, and IIQ-8 were respectively 0.87, 0.86, and 0.87 in the patient group, indicating good internal consistency. In the reference group, these were 0.91, 0.91, and 0.92 respectively, also indicating good internal consistency.

Reproducibility (Reliability and Agreement)

The retest assessments were completed on average 10 ± 25 days after baseline assessment. The ICCagreement was 0.84 for the UDI-6, and the domains ranged between 0.83 and 0.85, indicating good reliability (Table II). The ICCagreement of the IIQ-7 original, IIQ-7 adjusted, and IIQ-8 were 0.76, 0.77, and 0.75 respectively, also indicating good reliability.

The meanchange and the LOA in the test-retest period are also presented in Table II. The LOA are smaller than the MIC (see also results Interpretability), indicating good agreement.

Construct Validity

In 124 patients urodynamic data were available and classified according to urodynamic diagnosis. Sixteen of 124 patients were diagnosed with mixed incontinence and were classified into either SUI (n = 11) and DO (n = 4), while one was treated for mixed incontinence and therefore not included in this analysis. Table III presents the measure scores in 123 patients stratified into urodynamic diagnosis.

Table III. Construct Validity
 No urodynamic abnormality (n = 9)Detrusor overactivity (n = 34)Stress urinary incontinence (n = 59)Dysfunctional voiding (n = 9)Underactive Detrusor (n = 12)P-value#
  • Measure scores according to urodynamic diagnosis at baseline in mean ± SD.

  • All scores are transformed to a 0–100 point scale.

  • Higher scores indicate more symptom distress (UDI-6) or more impact on daily life (IIQ-SF).

  • Data in bold were used for pair wise comparisons of hypotheses using post hoc Games-Howell with mean difference significant at level *P < 0.05; **P < 0.01.

  • #

    Welch's F was used for comparing means.

UDI-6 (total score)39.5 ± 12.548.1 ± 13.148.2 ± 18.937.6 ± 17.943.5 ± 12.00.22
Irritative37.0 ± 23.2*68.6±23.151.6 ± 26.0*37.0 ± 23.2*55.5 ± 27.80.003
Stress50.0 ± 23.5*50.9 ± 24.2*66.9±22.850.9 ± 24.245.8 ± 21.4*0.006
Obstructive/discomfort31.5 ± 24.223.7 ± 22.825.0 ± 25.831.5 ± 24.229.1 ± 33.40.90
IIQ-7 original14.8±15.341.0 ± 22.4**38.8 ± 25.0**24.9 ± 19.333.7 ± 17.00.004
IIQ-7 adjusted15.3±15.241.7 ± 21.9**39.2 ± 25.8**25.5 ± 18.935.3 ± 16.30.003
IIQ-813.9±13.339.9 ± 22.2**37.5 ± 24.5**23.0 ± 17.933.0 ± 16.00.001

The distribution of the mean scores of the UDI-6 irritative domain, UDI-6 stress domain, IIQ-7 original, IIQ-7 adjusted, and IIQ-8, differed significantly between the urodynamic groups. Pairwise comparisons of predefined hypotheses on outcome resulted in 75% (3 out of 4) confirmed hypotheses indicating acceptable construct validity. Our confirmed hypotheses were:

  • The mean score of the irritative domain of the UDI-6 was significantly higher in the DO group than in patients with SUI (P = 0.02), dysfunctional voiding (P = 0.04), or without urodynamic abnormality (P = 0.02).
  • The mean score of the stress domain of the UDI-6 was significantly higher in the SUI group than in patients with DO (P = 0.02) or underactive detrusor (P < 0.05).
  • The mean scores of the IIQ-7 original, IIQ-7 adjusted, and IIQ-8 were significantly lower in the group of patients without urodynamic abnormality than in patients with SUI (P < 0.01) or DO (P < 0.01).

Responsiveness

The treatment work-up was individually determined for each patient by the treating physician after completion of the second round of questionnaires. Thus, 97 patients were treated and completed the third round of questionnaires on average 5.7 ± 1.3 months after baseline assessment. Treatment consisted of surgery (n = 71; 73%), conservative therapy (n = 16; 17%), or pharmaceutics (n = 10; 10%).

Table IV shows the meanchange between baseline and 6-month follow-up of each measure classified by the responses of the RAND-36-HTI. Since two patients (2%) reported their health was “much worse,” and 10 patients (11%) reported it was “a little worse” compared to 1-year ago, we combined those categories. The relationships between the meanchange of measure scores and the RAND-36-HTI as anchor were significantly linear (Table IV). The AUC for the UDI-6 was 0.81 (P < 0.001) indicating good responsiveness. The AUC for the IIQ-7 original was 0.65 (P = 0.02); for the IIQ-7 adjusted 0.70 (P = 0.004); and for the IIQ-8 0.66 (P = 0.02), indicating acceptable responsiveness.

Table IV. Responsiveness and Interpretability
 Number (%) (n = 97)UDI-6 (n = 80)IIQ-7 original (n = 81)IIQ-7 adjusted (n = 80)IIQ-8 (n = 80)
  • Data presented in number (%) or meanchange ± SDchange between baseline and 6-month follow-up.

  • Negative scores indicate improvement of urinary incontinence symptoms (UDI-6) and impact on daily life (IIQ-SF).

  • a

    Compared to 1-year ago, how would you rate your health in general now?

Health transition item (RAND-36)a
A little worse/much worse12 (12%)5.8 ± 19.7−2.8  ± 34.20.3 ± 30.5−1.8 ± 31.8
Same40 (41%)−7.8 ± 17.5−1.8 ± 22.2−1.7 ± 21.3−1.1 ± 21.8
A little better13 (13%)−19.3 ± 15.3−18.3 ± 20.6−20.1 ± 21.6−18.9 ± 21.0
Much better17 (18%)−27.6  ± 14.8−16.8 ± 28.6−21.2 ± 28.1−18.2 ± 27.0
Missing15 (16%)    
Pearson correlation coefficient r 0.540.250.340.29
P-value <0.0010.030.0020.01
Area under the ROC curve 0.810.650.700.66
P-value <0.0010.020.0040.02
Minimal important change −16.7−19.0−26.2−28.8
Sensitivity; specificity 0.76; 0.780.43; 0.840.41; 0.880.38; 0.90

Interpretability

The MIC was −16.7 for the UDI-6 with a sensitivity of 0.76 and specificity of 0.78. The MIC was −19.0 for the IIQ-7 original, −26.2 for the IIQ-7 adjusted, and −28.8 for the IIQ-8, with sensitivity ranging from 0.38 to 0.43 and specificity ranging from 0.84 to 0.90 (Table IV).

All MIC fell outside the LOA (Table II), indicating acceptable measurement errors which are smaller than the values for MIC.[15]

Floor and Ceiling Effects

We observed no floor or ceiling effects in the baseline data of our patient group. The lowest score was reported by at most 5% of respondents (IIQ-SF), and the highest score by at most 0.6% or respondents (IIQ-7 adjusted).

As expected in the reference group, no ceiling effects were observed, though floor effects were present in the UDI-6 (n = 76; 27%); IIQ-7 original (n = 212; 76%); IIQ-7 adjusted (n = 211; 76%); and IIQ-8 (n = 211; 76%).

DISCUSSION

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. ACKNOWLEDGMENTS
  9. APPENDIX
  10. REFERENCES

The objective of this study was to validate the UDI-6 and IIQ-7 in Dutch to provide a useful evaluation tool for the use in both men and women suffering from UI. In our population, the UDI-6 and IIQ-SF had good discriminant ability. The reproducibility (reliability and agreement) and construct validity of these measures were good, and both measures proved to be responsive and interpretable. By using the values of the MIC we determined, a decrease of ≥16.7 points in the UDI-6 score and a decrease of ≥19.0 points in the IIQ-7 original, ≥26.2 in the IIQ-7 adjusted, and ≥28.8 in the IIQ-8, can be considered as “clinically relevant” in patients undergoing any treatment for UI.

The internal consistency of the IIQ-SF in the patient group was good with Cronbach's alphas of 0.91 and 0.92. However, analysis of the UDI-6 showed moderate internal consistency of the total score and the domains, in both patient and reference group. Other validation studies found comparable Cronbach's alphas for the UDI-6.[20, 21] A possible explanation is the use of unconditional mean imputation[30] for measure scores, which is the scoring procedure as described by the developer.[6] This procedure allows imputing a score of one UDI-6 domain into another UDI-6 domain, for example, missing items of the irritative domain may be imputed by scores of the stress domain. As a consequence, domain scores might be “contaminated.” Another possible explanation of a moderate internal consistency of the UDI-6 was suggested by Franzen et al.[20] The UDI-6 contains items assessing symptoms of disease called “causal indicators.” These are called “causal indicators” since the occurrence of these symptoms can cause deterioration on HRQoL, but the impact of these symptoms on HRQoL may vary from patient to patient. Furthermore, the patient's perception of the severity of their symptoms may be influenced by other factors and concomitant symptoms.[34] The IIQ-SF contains items measuring effects of UI on HRQoL, which reflect the level of HRQoL, and are regarded as “effect indicators.”[34] Scales based on causal indicators are more heterogeneous than scales based on effect indicators. For example, a patient with family problems may perceive UI as having a greater effect upon their HRQoL than usual. Since psychometric analysis is based on the assumption that all items are effect indicators reflecting the same latent HRQoL construct, the properties of causal items (UDI-6) affect reliability measures such as Cronbach's alpha leading to a lower internal consistency.[34]

We recognize several limitations to this study. First, we recruited patient data from a tertiary urology and gynecology centre. In this setting we expect more patients with severe symptoms of UI, and therefore the extent to which the results of this study can be generalized to other settings may be limited. Second, we did not validate the measures separately for men and women. A comparison between the two genders is desirable since health seeking behaviour and bother from lower urinary tract symptoms (LUTS)[35] and UI[36, 37] differs significantly between men and women. Twenty-five present of women were bothered by the presence of LUTS compared to 18% men, with equal prevalence of treatment seeking of 28% and 24% respectively, however, in the absence of evaluation of the effect of incontinence.[35] In this hospital-based survey, men were more likely than women to seek treatment as age and severity of LUTS increased.[35] When UI was taken into account in a community based survey, 45% of women versus 22% of men suffering from incontinence had sought medical care for their problem.[36] However, in a different community based study, men with incontinence reported a higher rate of healthcare seeking of 56% compared to 46% of women.[37] Even though we did not intend to compare sexes and their urological functioning as mirrored by the questionnaires, we can highly recommend this for future research. This study was not designed for this purpose as we would have taken gender distribution into account in our sampling procedure.

The strength of this study is the study design, which enabled us to address almost all quality criteria for measurement properties of Terwee et al.[31] An exception was the criterion validity, which requires the availability of a gold standard. Since there are no gold standards of symptoms and bother of UI, we were not able to test this property. Our study design also allowed us to test the UDI-6 and IIQ-SF in male patients. As suggested in an earlier study,[23] we revised the gender-specific items of the IIQ-7. Hence, we studied the original IIQ-7 along with two variants: the IIQ-7 “adjusted” and the IIQ-8. The psychometric properties of these variants of IIQ-SF were equally good. Therefore we would like to recommend using the “original” IIQ-7 in women as well as in men, to allow better international comparison of research and treatment outcome data.

CONCLUSION

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. ACKNOWLEDGMENTS
  9. APPENDIX
  10. REFERENCES

In conclusion, the UDI-6 and IIQ-7 in Dutch are reliable, valid and responsive instruments, and therefore suitable for assessing symptom distress and HRQoL of UI in both men and women, as well as evaluating outcome of treatment.

ACKNOWLEDGMENTS

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. ACKNOWLEDGMENTS
  9. APPENDIX
  10. REFERENCES

Thanks to John S. Uebersax for giving permission to validate and his advice how to validate these measures in male patients; to Pim J. van Leeuwen for his contribution to the design of this study; to Christa A. Lips and Ben Wasscher for their translations of the questionnaires; to Caroline B. Terwee for her methodology advice; to Gerard J.J.M. Borsboom for his statistical assistance; and especially to all the patients who gave their time and effort to participate. This study was funded by the Stichting Urologisch Wetenschappelijk Onderzoek (SUWO, Rotterdam, the Netherlands) to make reference data collection possible. SUWO was not involved in the design, conduct, interpretation, or analysis of the study or review or approval of the manuscript.

APPENDIX

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. ACKNOWLEDGMENTS
  9. APPENDIX
  10. REFERENCES

How much are you bothered by (check one answer for each question):

  1. Frequent urination
  2. Leakage related to feeling of urgency
  3. Leakage related to activity, coughing, or sneezing
  4. Small amounts of leakage (drops)
  5. Difficulty emptying bladder
  6. Pain or discomfort in lower abdominal or genital area

Item response levels are: 0=not at all, 1=slightly, 2=moderately, and 3=greatly.

Incontinence Impact Questionnaire Short Form (IIQ-SF) (English version)1

Some people find that accidental urine loss may affect their activities, relationships, and feelings. The questions below refer to areas in your life which may have been influenced or changed by your problem.

For each question, check the response that best describe how much your activities, relationships, and feelings are being affected by urine leakage.

Has urine leakage affected your (check one answer for each question):

  1. Ability to do household chores (cooking, housecleaning, laundry)?
  2. Physical recreation such as walking, swimming, or other exercise?
  3. Entertainment activities (movies, concerts, etc.)?
  4. Ability to travel by car or bus more than 30 min from home?
  5. Employment (work) outside the home?
  6. Participation in social activities outside your home?
  7. Emotional health (nervousness, depression, etc.)?
  8. Feeling frustrated?

Item response levels are: 0=not at all, 1=slightly, 2=moderately, and 3=greatly.

UDI-6: Urogenital Klachten Lijst (Dutch version)

In hoeverre heeft u van het volgende last (kruis één antwoord per vraag aan):

  1. Vaak plassen
  2. Urineverlies dat verband houdt met het gevoel van aandrang
  3. Urineverlies dat verband houdt met lichamelijke activiteit, hoesten of niezen
  4. Kleine hoeveelheden urineverlies (druppels)
  5. Moeite uw blaas te legen
  6. Pijn of ongemak in de onderbuik of rond het kruis

Antwoordmogelijkheden: 0=helemaal niet, 1=een beetje, 2=redelijk wat, 3=zeer veel.

IIQ-SF: Vragenlijst Impact van Incontinentie (Dutch version)

Sommige mensen vinden dat ongewild urineverlies invloed kan hebben op hun activiteiten, relaties en gevoelens. De onderstaande vragen gaan over facetten van uw leven die misschien zijn beïnvloed of veranderd door uw probleem.

Kruis bij elke vraag het antwoord aan, dat het best omschrijft in hoeverre uw activiteiten, relaties en gevoelens zijn beïnvloed door urineverlies.

Heeft urineverlies invloed gehad op uw (kruis één antwoord per vraag aan):

  1. vermogen om huishoudelijke taken te doen (koken, schoonmaken, de was doen)?
  2. lichamelijke activiteiten zoals bijv. wandelen of zwemmen of andere oefeningen?
  3. uitgaansactiviteiten (films, concerten, etc.)?
  4. vermogen om langer dan 30 minuten van uw huis met de auto of bus te reizen?
  5. baan (werk) buitenshuis?
  6. deelname aan sociale activiteiten buitenshuis?
  7. emotionele gezondheid (nervositeit, depressie, etc.)?
  8. gevoel van frustratie?

Antwoordmogelijkheden: 0=helemaal niet, 1=een beetje, 2=redelijk wat, 3=zeer veel.

REFERENCES

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. ACKNOWLEDGMENTS
  9. APPENDIX
  10. REFERENCES
  • 1
    Milsom I. Epidemiology of urinary and faecal incontinence and pelvic organ prolapse. In: Abrams P, Cardozo L, Khoury S, Wein A, editors. 4th International Consultation on Incontinence. 4th ed. Paris: Health Publication Ltd; 2009.
  • 2
    Patrick DL, Deyo RA. Generic and disease-specific measures in assessing health status and quality of life. Med Care 1989; 27:S217S32.
  • 3
    Korfage IJ, Hak T, de Koning HJ. et al. Patients' perceptions of the side-effects of prostate cancer treatment—A qualitative interview study. Soc Sci Med 2006; 63:9119.
  • 4
    Shumaker SA, Wyman JF, Uebersax JS. et al. Health-related quality of life measures for women with urinary incontinence: The Incontinence Impact Questionnaire and the Urogenital Distress Inventory. Continence Program in Women (CPW) Research Group. Qual Life Res 1994; 3:291306.
  • 5
    Wyman JF, Harkins SW, Choi SC. et al. Psychosocial impact of urinary incontinence in women. Obstet Gynecol 1987; 70:37881.
  • 6
    Uebersax JS, Wyman JF, Shumaker SA. et al. Short forms to assess life quality and symptom distress for urinary incontinence in women: The Incontinence Impact Questionnaire and the Urogenital Distress Inventory. Continence Program for Women Research Group. Neurourol Urodyn 1995; 14:1319.
  • 7
    Staskin D, Kelleher C. Incontinence, Patient-Reported Outcome Assessment. 4th International Consultation on Incontinence; July 5–8, 2008, 2009; Paris.
  • 8
    Guillemin F, Bombardier C, Beaton D. Cross-cultural adaptation of health-related quality of life measures: Literature review and proposed guidelines. J Clin Epidemiol 1993; 46:141732.
  • 9
    van der Vaart CH, de Leeuw JR, Roovers JP. et al. Measuring health-related quality of life in women with urogenital dysfunction: The urogenital distress inventory and incontinence impact questionnaire revisited. Neurourol Urodyn 2003; 22:97104.
  • 10
    van der Vaart CH, de Leeuw JR, Roovers JP. et al. The influence of urinary incontinence on quality of life of community-dwelling, 45–70 year old Dutch women. De invloed van urine-incontinentie op de kwaliteit van leven bij thuiswonende Nederlandse vrouwen van 45–70 jaar. Ned Tijdschr Geneeskd 2000; 144:8947.
  • 11
    Labrie J, van der Graaf Y, Buskens E. et al. Protocol for physiotherapy or TVT Randomised Efficacy Trial (PORTRET): A multicentre randomised controlled trial to assess the cost-effectiveness of the tension free vaginal tape versus pelvic floor muscle training in women with symptomatic moderate to severe stress urinary incontinence. BMC Womens Health 2009; 9:24.
  • 12
    van Leijsen SA, Kluivers KB, Mol BW. et al. Protocol for the value of urodynamics prior to stress incontinence surgery (VUSIS) study: A multicenter randomized controlled trial to assess the cost effectiveness of urodynamics in women with symptoms of stress urinary incontinence in whom surgical treatment is considered. BMC Womens Health 2009; 9:22.
  • 13
    Vervest HA, Bisseling TM, Heintz AP. et al. The prevalence of voiding difficulty after TVT, its impact on quality of life, and related risk factors. Int Urogynecol J Pelvic Floor Dysfunct 2007; 18:17382.
  • 14
    van Leijsen SA, Kluivers KB, Mol BW. et al. Can preoperative urodynamic investigation be omitted in women with stress urinary incontinence? A non-inferiority randomized controlled trial. Neurourol Urodyn 2012; 31:111823.
  • 15
    de Vet HCW, Terwee CB, Mokkink LB. et al. Measurement in medicine. 1st ed. Cambridge: Cambridge University Press; 2011.
  • 16
    Altaweel W, Seyam R, Mokhtar A. et al. Arabic validation of the short form of Urogenital Distress Inventory (UDI-6) questionnaire. Neurourol Urodyn 2009; 28:3304.
  • 17
    Cam C, Sakalli M, Ay P. et al. Validation of the short forms of the incontinence impact questionnaire (IIQ-7) and the urogenital distress inventory (UDI-6) in a Turkish population. Neurourol Urodyn 2007; 26:12933.
  • 18
    Chan SS, Choy KW, Lee BP. et al. Chinese validation of urogenital distress inventory and incontinence impact questionnaire short form. Int Urogynecol J 2010; 21:80712.
  • 19
    El-Azab AS, Mascha EJ. Arabic validation of the urogenital distress inventory and adapted incontinence impact questionnaires—Short forms. Neurourol Urodyn 2009; 28:339.
  • 20
    Franzen K, Johansson JE, Karlsson J. et al. Validation of the Swedish version of the incontinence impact questionnaire and the urogenital distress inventory. Acta Obstet Gynecol Scand 2013; 92:55561.
  • 21
    Ruiz de Vinaspre Hernandez R, Tomas Aznar C, Rubio Aranda E. Validation of the Spanish version of the short forms of the Urogenital Distress Inventory (UDI-6) and the Incontinence Impact Questionnaire (IIQ-7) in pregnant women. Validacion de la version espanola de las formas cortas del Urogenital Distress Inventory (UDI-6) y del Incontinence Impact Questionnaire (IIQ-7) en mujeres embarazadas. Gac Sanit 2011. 25:37984.
  • 22
    Zhu L, Yu SJ, Lang JH. et al. Validation of incontinence impact questionnaire short form in Chinese population. Zhonghua Fu Chan Ke Za Zhi 2011; 46:5059.
  • 23
    Moore KN, Jensen L. Testing of the incontinence impact questionnaire (IIQ-7) with men after radical prostatectomy. J Wound Ostomy Continence Nurs 2000; 27:30412.
  • 24
    Abrams P, Cardozo L, Fall M. et al. The standardisation of terminology of lower urinary tract function: Report from the Standardisation Sub-committee of the International Continence Society. Neurourol Urodyn 2002; 21:16778.
  • 25
    Zee vdK, Sanderman R. Het meten van de algemene gezondheidstoestand met de RAND-36. Groningen: Een handleiding.
  • 26
    Crosby RD, Kolotkin RL, Williams GR. Defining clinically meaningful change in health-related quality of life. J Clin Epidemiol 2003; 56:395407.
  • 27
    Abrams P, Andersson KE, Birder L. et al. Fourth International Consultation on Incontinence Recommendations of the International Scientific Committee: Evaluation and treatment of urinary incontinence, pelvic organ prolapse, and fecal incontinence. Neurourol Urodyn 2010; 29:21340.
  • 28
    Schafer W, Abrams P, Liao L. et al. Good urodynamic practices: Uroflowmetry, filling cystometry, and pressure-flow studies. Neurourol Urodyn 2002; 21:26174.
  • 29
    Flycatcher Academic Support Centre. http://www.flycatcher.eu/nld/academic-support/, 2012.
  • 30
    Little RJA, Rubin DB. Statistical analysis with missing data. 2nd [rev] ed. Hoboken N.J.: Wiley Interscience; 2002.
  • 31
    Terwee CB, Bot SD, de Boer MR. et al. Quality criteria were proposed for measurement properties of health status questionnaires. J Clin Epidemiol 2007; 60:3442.
  • 32
    McGraw K, Wong S. Forming inferences about some intraclass correlation coefficients. Psychol Methods 1996; 1:3046.
  • 33
    Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1:30710.
  • 34
    Fayers PM, Hand DJ, Bjordal K. et al. Causal indicators in quality of life research. Qual Life Res 1997; 6:393406.
  • 35
    Apostolidis A, Kirana PS, Chiu G. et al. Gender and age differences in the perception of bother and health care seeking for lower urinary tract symptoms: Results from the hospitalised and outpatients' profile and expectations study. Eur Urol 2009; 56:93747.
  • 36
    Harris SS, Link CL, Tennstedt SL. et al. Care seeking and treatment for urinary incontinence in a diverse population. J Urol 2007; 177:6804.
  • 37
    Li Y, Cai X, Glance LG. et al. Gender differences in healthcare-seeking behavior for urinary incontinence and the impact of socioeconomic status: A study of the Medicare managed care population. Med Care 2007; 45:111622.
Note
  1. 1

    Adapted from: Refs[4, 6].