Dr K. H. Moore, St George Hospital, First Floor, Pitney Clinical Sciences Building, Kogarah, Sydney, New South Wales 2217, Australia.
The values for ‘mild’, ‘moderate’ and ‘severe’ urinary incontinence have not been determined for the 24-hour pad test. To define these values, a prospective observational study was performed on 110 women with the primary symptom of urinary incontinence. Consenting women performed two 1-hour pad tests one week apart, and seven 24-hour pad tests for seven consecutive days. The 1-hour pad test definitions for mild, moderate and severe were translated to centiles, and used to categorise the 24-hour test values. This revealed that the range for ‘mild incontinence’ was between 1.3 and 20 g, ‘moderate incontinence’ ranged from 21 to 74 g, and ‘severe incontinence’ was defined as 75 g or more in 24 hours. Severity of leakage was analysed in relation to urodynamic diagnosis, age, parity and pelvic floor muscle strength. Increasing severity was associated with increasing age and parity. Women with detrusor overactivity were most likely to have severe leakage. In conclusion, this study defines the three grades of severity for the 24-hour pad test, which may help to guide patients' choice between conservative and surgical treatment and is useful for stratified randomisation of controlled trial participants.
Objective assessment of severity is an important part of the clinical evaluation of patients with urinary incontinence. At present, only the 1-hour pad test has been used to define the terms ‘mild’ (2–10 mL/h), ‘moderate’ (11–50 mL/h) and ‘severe’ (>50 mL/h) quantitatively.1 Knowledge of the degree of severity may allow clinicians to offer a conservative regime to those with mild incontinence, or recommend surgical treatment to those with severe leakage.2 Furthermore, definitions of mild, moderate and severe may be helpful when performing controlled trials, if stratified randomisation with respect to severity is required.
The optimal outcome measure in the field of urinary incontinence has not been agreed upon, but the test–retest reliability of the 1-hour pad test has come under considerable criticism.3 The 24-hour pad test appears to give a more realistic appraisal of the typical urine loss during ordinary daily activity, but the values for mild, moderate and severe have not yet been defined for this test. Recently, the upper limit of ‘normal’ for the 24-hour pad test has been redefined for continent women as being 1.3 g.4 With this in mind, we aimed to establish the categories of mild, moderate and severe incontinence, and to use the most recent definition of normal as the lower limit of ‘mild incontinence’.
A consecutive series of 110 women attending a tertiary urogynaecological unit with the complaint of urinary incontinence were recruited. Inclusion criteria were community dwelling status, ability to speak and read English and sufficient mobility and dexterity to perform both the 1-hour and 24-hour pad tests. Those with a primary complaint of prolapse, recurrent cystitis and voiding difficulty were excluded from the study. All women completed the study either just before or soon after urodynamic diagnosis, but all were recruited prior to the initiation of any treatment. The study was conducted with the approval of the ethics committee, and informed consent was obtained for each patient.
Participants were asked to perform two standardised 1-hour pad tests one week apart. The 1-hour pad test was performed according to ICS recommendations, with a bladder volume greater than 200 mL (measured using a BladderScan). This bladder volume was achieved by natural diuresis following the ingestion of 500 mL of water. The participants performed manoeuvres and exercises during the 1-hour period while wearing a pre-weighed absorbent sanitary pad (Tena Normal), which was again weighed at the completion of the provocation exercises for calculation of urine loss.
Seven 24-hour pad tests were performed during the week following the second 1-hour pad test. Participants were given seven resealable plastic bags (one for each 24-hour period) and enough pre-weighed pads for seven days. All pads used in any 24-hour period were placed in the sealable plastic bags and returned at the completion of seven consecutive 24-hour periods. The pads were weighed and the urinary leakage for each 24-hour period was calculated. The weights of backing strips and wrappers were also factored in, if not included in the returned sealable plastic bags. The scales used to weigh pads were digital kitchen scales that were accurate to 1 g.
In addition to pad testing, all women were asked to complete a 24-hour frequency volume chart indicating volumes and times of fluid intake and output and activities conducted through each of the seven days.
Following the completion of all pad tests, the study patients underwent usual care at the unit, with history, clinical examination (including the Oxford pelvic floor muscle grading system)5 and urodynamic assessment. Urodynamic diagnoses were classified into four types: urodynamic stress incontinence, detrusor overactivity incontinence, mixed incontinence (urodynamic stress incontinence + detrusor overactivity incontinence) and miscellaneous diagnoses (including sensory urgency and voiding dysfunction).
The original authors who defined mild, moderate and severe incontinence for the 1-hour test, did so by dividing the volume leaked in 1 hour by the ultrasound-determined volume remaining in the bladder at the end of the test plus the volume leaked, then stratifying the resulting percentage leak into equal thirds. Such calculation is not possible for the 24-hour test as total bladder volume is too difficult to measure over 24 hours. Thus, to determine mild, moderate and severe for the 24-hour pad test, the two sets of 1-hour pad tests for all patients were examined. These two sets of results were pooled, regardless of diagnosis, yielding 219 data. The percentage of the data which was classified as mild, moderate or severe on the 1-hour tests was noted, as these percentages were used to divide the 24-hour tests into the same three categories. The pooled data set of 758 twenty-four-hour pad tests were sorted in ascending order of magnitude, and then divided into mild, moderate and severe based on the percentages derived from the 1-hour test.
The relationship between our definition of ‘mild’, ‘moderate’ and ‘severe’ and factors known to effect severity, such as age, parity, pelvic floor strength and urodynamic diagnosis, were displayed using 100% stacked bar charts (Fig. 1). Where appropriate, Spearman correlation and χ2 tests were used to test these relationships.
Statistical analysis was performed using SPSS version 11 statistical software. Results were considered significant at the 5% level (i.e. P < 0.05).
One hundred and ten women aged 26 to 85 years (median age 57 years, interquartile range [IQR] 48–68) consented to participate in the study. One hundred and four patients completed both 1-hour pad tests and all seven 24-hour pad tests. The remainder completed between four and six 24-hour pad tests (5) or only one 1-hour pad test (1). This yielded a total of 758 twenty-four-hour pad tests out of a possible 770. The women had a median parity of 3 (IQR 2–3) and a median pelvic floor muscle grade 2 out of 5 (IQR 2–3). Urodynamic investigation subsequently revealed urodynamic stress incontinence in 67 (65%), mixed incontinence (urodynamic stress incontinence and detrusor overactivity) in 30 (29%), detrusor overactivity incontinence in 6 (6%). Miscellaneous diagnoses included sensory urgency (n= 3), voiding difficulty (n= 2) or a mixture thereof (n= 2) which were excluded from the subset analysis by diagnosis, owing to the small numbers.
The median pad weight gain during all (n= 219) 1-hour pad tests was 6 g (IQR 0–28 g). Of these, 41.3% revealed pad loss of ≤2 g (i.e. continent), 17% revealed pad loss between 2 and 10 g (mild), 22.4% of subjects lost 11 to 50 g (moderate) and 15% had losses greater than 50 g (severe).1 Considering only the first 1-hour pad test, 43.7% were ‘dry’ (i.e. results were ≤2 g). This was similar to the 38.8% rate for ‘dry’ pad tests on the second occasion (χ2= 0.31, P= 0.57). Thirty-three (32%) of those women tested had ≤2 g pad loss on both pad tests in keeping with our previous results. For the first of the two 1-hour pad tests, the median weight gain was 4 g (IQR 0–14 g), while the second test revealed a median gain of 6 g (IQR 0–28). Because all women attended the unit with a main complaint of incontinence, those with a ‘dry’ pad test were allocated to the lower end of the ‘mild’ category for the purposes of assigning centile cutoff values between mild and moderate.
A total of 758 twenty-four-hour pad tests were undertaken by 110 women. The median pad test weight gain in a 24-hour period was 12.0 g (IQR 4.0–51.3). Using the recently determined upper normal limit of 1.3 g4 (which in this study was rounded to 1 g because of the use of less accurate scales), a total of 93 (13%) 24-hour pad tests had a weight gain of ≤1 g. Looking at the sensitivity of the 24-hour tests, only one patient had all seven 24-hour pad tests weighing ≤1 g. Of the remaining participants, only eight (7.7%) had two pad tests weighing ≤1 g, six (5.5%) had three pad tests weighing ≤1 g, four (3.8%) had four pad tests weighing ≤1 g, two (1.9%) had five ‘continent’ tests.
Of the 219 one-hour pad tests, mild leakage was observed in 57.7% of cases and moderate leakage was seen in a further 23.2% of cases. Therefore, these centiles were applied to the 24-hour data set. Because ‘continent’ on the 1-hour pad test is defined as pad weight gain of 2 g or less, a separate analysis was performed in which these values were excluded from the derivation of mild, moderate and severe. This repeat analysis showed little change in centiles (52.0% and 26.3%).
We examined the total pooled 24-hour data (n= 758) and found that the centile cutoff levels for the 1-hour results corresponding to mild/moderate and moderate/severe (52.0% and 78.3%) equated to 20 and 74 mL on the 24-hour pad test.
As shown in Fig. 1, as age increased, the proportion of severe cases increased (r= 0.147, P≤ 0.001). Severity of leakage rose sharply with rising numbers of vaginal deliveries (r= 0.356, P≤ 0.001). Poor pelvic floor muscle strength bore little relationship to severity of pad test leakage (r= 0.064, P= 0.08). Patients with pure detrusor overactivity incontinence had the most severe leakage (χ2= 79.6, P≤ 0.001).
Before the commencement of any treatment of urinary incontinence, two fundamental issues must be addressed, firstly, what type of incontinence does the patient have, and secondly, what is the severity of the disease. Clinical examination and urodynamic investigation will differentiate the various types of incontinence but quantification of severity has previously only been available for the 1-hour pad test.
Perineal pad testing falls into two broad categories, namely, laboratory-based short tests and longer home-based testing. Laboratory-based pad testing has been criticised as being inherently artificial in nature, because of supra-physiological bladder filling volumes, provocation events that may not reflect those causing leakage in daily life, and embarrassment.3 The reproducibility of the laboratory-based test is improved by standardising the bladder volume using catheterisation retrograde fill, but this increases the overall invasiveness of the procedure. As explained the categories of mild, moderate and severe have been firmly established for this test.1
Home-based pad testing allows the patient to undergo testing in a familiar environment, while performing those provocative manoeuvres that cause urinary leakage. The 24-hour pad test has been criticised because of the undefined optimal duration of testing (i.e. 24 vs 48 vs 72 hours) and the fact that patient compliance in returning the wet pads for weighing is less than 100%.6 To date, the longer pad test has also suffered from a lack of definition of the categories of mild, moderate and severe.
While the lower limit of ‘mild’ (or the upper limit of ‘normal’) has recently been determined as being 1.3 g, this measurement was performed using scales accurate to 0.1 g in a study that was undertaken after we had commenced the current project. As the scales used in the current study were accurate to 1 g, we rounded 1.3 g off to 1 g, which became the demarcation between normal and mild incontinence for the 24-hour pad test. The definition of ‘continent’ (2 g) for the 1-hour pad test was also made by using relatively inaccurate kitchen scales. Because the centile cutoffs for mild and moderate were not markedly different when the 1 g definition of continence was used (52%vs 57.7% and 26.3%vs 23.2%, respectively), we opted to maintain the derivation of the categories using 1 g as the lower limit of normal.
In the present study, we have used the pooled data of two 1-hour pad tests to define the percentage of our 109 women who were classified as ‘mild, moderate, severe’, then applied these centile cutoffs to the 751 home-based 24-hour tests performed by these women. Thus, we have derived the upper limits of mild incontinence as 20 mL, and the lower limits of severe incontinence as 75 mL.
We suggest that the values for grades of severity for leakage on 24-hour pad testing should be 1.3–19 g < (<20 g) for mild incontinence, 20–74 g for moderate incontinence and 75 g or greater for severe incontinence. We hope that these findings will encourage greater reliance upon the 24-hour pad test to help determine choice of conservative versus surgical therapy in mild versus severe cases, to aid in stratified randomisation of controlled trials and to clarify definition of cure, partial response and failure in outcome studies.