In order to treat patients with urinary incontinence (UI), a baseline and post-treatment comprehensive assessment is required including first of all an objective quantification of the degree of involuntary leakage [1]. The evaluation of the degree of UI is a key aspect of the UI management with containment products in order to personalize the choice of a well-fitted product as much as possible. We reported that after adapting the products prescription to the objectively measured leakage volume, the appropriateness of pads use was significantly improved [2]. Similarly, an accurate measure of UI severity is of utmost importance in order to choose the best pharmacological or surgical treatment and to compare the outcomes of different treatment strategies.
Pad test is recognized as a robust and reproducible tool for the measurement of the degree of UI [2]. On the other hands, the number of pads used per day is widely used as an estimation of UI severity and has shown to be a reliable measure of urine leakage volume in an ambulatory setting [3]. While pad tests are somewhat cumbersome and time-consuming, pad count recorded by the patients/carers for 24 hours or some days is definitely a more streamlined method.
The objective of this study was to determine in a very large population of community-dwelling incontinent patients, the accuracy of pad count as measure of the degree of incontinence, using data from a multicenter, large-scale, pad test study.

The “Determining the Individual Appropriateness of Pads Provision and Enhancing its Realization” (DIAPPER) pad test study has been the largest published study evaluating the appropriateness of continence products prescription [2]. It has been an Industry-supported research performed in five continence services of Italy. In the present study, data obtained from the first cross-sectional part of the study were used for the analyses. 
Briefly, patients suffering from UI and provided with absorbent products were included from 01/2012 to 03/2016. All patients or their caregivers have been invited to perform a 48-hour pad test in their usual home environment and to fill in a diary with detailed information on pad usage, including number of changes. Patients/cares not willing to participate, incomplete filled diary, and/or not adequately performed pad test were criteria for exclusion.
At visit 1, patients/carers were carefully instructed on how to perform the pad test. Patients had to complete the pad test performing usual daily activities, and using their usual pads and changing them as they would regularly. At visit 2, the diary were analysed by the continence professionals and the following data were recorded: number, design and size of used products, wearing time (WT) and net pad weight gain (PWG) of each product, total amount of urine lost in 48 hours (48PWG). Pads size was based on the absorption level they are intended for, defined as follows: up to 200 g/pad (light UI); from 200 to 350 g/pad (moderate UI); 350 to 500 g/pad (severe UI); >500 g/pad (very severe UI). 
Descriptive statistics for each of the three measures of UI (PWG, 48PWG, 48-hour pad count) were computed. Pearson’s correlation coefficients between all variables were calculated. To investigate patients’ characteristics influencing pad count, univariate and multivariate logistic regression analyses were conducted using LOGIT and GLM model. Included variables have been age, gender, body weight, waist circumference, health district of residence, type of main activity during the study days (walking, sitting, lying), level of autonomy (autonomous or assisted) and mobility (ambulatory, partially ambulatory, bedridden), skin health status (healthy, inflamed, ulcerated), use of diuretics (yes, no), number, design (rectangular, shaped, brief, pull-on, bed pad) and size of used products, mean wearing time (WT) and mean PWG, 48PWG. Differences were considered to be significant when P < 0.05.

Out of 19,675 screened patients, 14,493 patients/cares (94.2%) completed correctly the pad test and were included in the final analysis, using overall during the study days 98,362 continence products. Males and females represented the 26% and 74%, respectively, of the total population. Mean age was 78±15 years (79±14 for males, 74±19 for females). 
Mean 48-hour pad count was 6.8 (±2.7, range 1-12) and mean PWG and 48PWG were 290 g (±243, range 1-2144) and 1966 g (±1569, range 5-16,250), respectively. Male patients showed higher 48PWG and PWG, but lower 48-hour pad count compared with females (p<0.0001 for all comparisons). 
The Scatterplot of 48-hour pad count versus 48PWG is shown in Figure 1. 48-hour pad count showed only a weak positive correlation with 48PWG. The linear correlations for the whole population showed an R2 value of 0.12. The correlations for males and females showed an R2 value of 0.18 and 0.11, respectively. A weak negative correlation (-0.16; p<0.01) was observed between PWG and 48-hour pad count. Very week correlations were found between the different UI measures and age.
Higher correlations between 48PWG and 48-hour pad count were found among patients using <6 pads/48-hour (R2 = 0.14) compared with those using >6 pads/48-hour (R2 = 0.02; p<0.01), although correlation coefficients remained low in both range groups. 
At multivariate logistic regression analyses, an independent association between pad count and both 48PWG and PWG was observed, although the relationships were not linear. A statistically significant negative association between pad size and pad count was observed, with patients using products for severe and very severe UI showing a 45% and 39% lower propensity to use more pads than those using products for light UI (p<0.0001), respectively. With regard to product design, patients using shaped and rectangular products had 34% and 40% higher propensity to use more pads than those using briefs (p<0.0001), respectively. 
Lower level of activity gave a higher propensity to use more pads, and the health district of residence as well was a weak but independent predictor of pad count.

There was a very weak correlation between the pad count and the degree of UI as measured with pad test. Pad count only accounted for 13% of the variability of UI volume (18% and 11% in males and females, respectively).
The relationship between pad count and pad weight was not linear and the higher the pad count, the lower the correlation. Thus a stepwise increase of the 48PWG was observed particularly in the range of 48-hour pad count from one to six with the highest correlation coefficient observed for male patients using <6 pads/48-hour (R2 = 0.20). Consequently, pad count appears still more unhelpful in distinguishing patients using a number of pads higher than six. 
The complexity of the relationship between pad count and pad weight is further demonstrated by the impact that variables other than leakage volumes had on the propensity (odds) to use more pads.

This study confirmed on a very large observational base that pad count is a poor measure of the degree of UI having a very modest correlation with the real leakage volume as objectively measured by the pad weighing test. Pad count only measured 13% of the variance of UI volume. Consequently, pad count should not be used instead of the pad test as an objective measure of UI severity when an accurate evaluation is required for research or clinical purposes.

Castro-Diaz, Robinson D, et al. Initial assessment of urinary incontinence in adult male and female patients. In: Abrams P, Cardozo L, Wagg A, Wein A. (eds) Incontinence 6th International Consultation on Incontinence, Tokyo, ICUD-ICS 2016
Sacco E, Bientinesi R, Gandi C, Palermo G, Pierconti F, Bassi P. Objectively improving appropriateness of absorbent products provision to patients with urinary incontinence: The DIAPPER study. Neurourol Urodyn. 2018 Jan;37(1):485-495
Groutz A, Blaivas JG, Chaikin DC, Resnick NM, Engleman K, Anzalone D, Bryzineski B, Wein AJ. Noninvasive outcome measures of urinary incontinence and lower urinary tract symptoms: A multicenter study of micturition diary and pad tests. J Urol 2000; 164: 698–701