Rodney Brooks

Joint Continuum Regression for Multiple Predictands

Abstract This article generalizes continuum regression (CR) in the hope that regressors “jointly constructed” for several predictands might improve on the separate prediction of individual predictands. The generalization developed is a mixture of principal components regression and de Jongs modification of partial least squares for multiple predictands. The balance of ingredients can be chosen by cross-validation, as can the number of regressors constructed. The new method has been tested on real and simulated data. The indications are that conditions for the superiority of the joint approach may be rare in practice.

Coated catheters for intermittent catheterization: smooth or sticky?

Objective To evaluate the current range of hydrophilic‐coated catheters for intermittent self‐catheterization, focusing on the adherence of the catheter to the urethral mucosa at the end of catheterization.

Sheaths for urinary incontinence: a randomized crossover trial.

Objective To evaluate the full range of self‐adhesive continence sheaths for men available in the UK and thus provide clinicians and consumers with a basis for product selection.

Predicting the leakage performance of small bodyworn disposable incontinence pads using laboratory tests

An international multi-centre project (The ISO Pad Leakage Project) was conducted to study the leakage performance of large bodyworn incontinence pads for heavily incontinent users and to create international standards for measuring their absorption capacities in the laboratory. This was achieved by recruiting 13 user test centres through which over 100 incontinent subjects tested each of six different products (each in three different sizes) for a period of about a week to a common protocol. Over 10,000 used pads were collected and weighed and the severity of leakage from each of them recorded. Correlations were sought between these data and the results from some 50 technical tests performed in a total of 16 technical test centres in order to discover the impact of different technical parameters on clinical pad performance. It was found that at low urine weights (less than 50 g, say) pad shaping was the most important predictor of pad leakage performance: shaped pads leaked less. With increasing urine weight, absorption capacity and absorption time increased in importance until at 350 g of urine these two parameters and shaping were of about equal significance: shaped pads with high absorption capacity and fast absorption time leaked least. A second series of analyses identified two absorption capacity tests which produced data correlating well with the overall leakage performance of pads, considering all urine weights together. Both tests were checked for repeatability (precision within laboratories) and reproducibility (precision between laboratories) and have been written up as working draft standards.(ABSTRACT TRUNCATED AT 250 WORDS)

How well does ISO 11948-1 (the Rothwell method) for measuring the absorption capacity of incontinence pads in the laboratory correlate with clinical pad performance

The ability of ISO 11948-1 (the Rothwell method) to predict the leakage performance of disposable bodyworn pads for heavy urinary incontinence was investigated by measuring correlations between models based on clinical evaluations of 138 diapers and inserts (the two major design categories), and technical models based on their Rothwell absorption capacities and design features. Correlations were poorer than in the original 1993 study for the standard (r < or =0.87 compared with r < or =0.95), but still strong enough to help with purchasing choices. For a given Rothwell capacity, the leakage performance of diapers was far superior to inserts; for example, diapers containing 450 and 300 g of urine performed, as well as inserts containing 300 and 100 g, respectively. No evidence was found for any other design feature having a significant impact on leakage performance. The coefficient of variation for Rothwell capacity (a measure of product consistency) had significant impact on the leakage performance of diapers, but not inserts. The probability of diapers with the poorest consistency leaking exceeded that for the best by about 10 percentage points. Similarly, diapers were about 10 percentage points more likely to leak when used at night than during the day. Differences between day-time and night-time use of inserts were not studied.

Disposable bedpads for incontinence: predicting their clinical leakage properties using laboratory tests.

A multi-centre project has been run to identify laboratory tests capable of predicting the leakage performance of disposable incontinence bedpads. Each of 95 subjects tested each of six products for a week in turn and reported whether or not they and/or their carers found the leakage performance of each product acceptable. In addition, carers noted the severity with which individual used bedpads had leaked so that, when they had been weighed, their leakage performance could be determined as a function of urine weight. These clinical data were compared with results from the 16 different laboratory tests used routinely for bedpad evaluation in three hospital laboratories. Each test was evaluated by seeing how well the data it yielded correlated with the clinical test data. No individual test was very successful at predicting the performance of bedpads when used as sole protection but a combination of an absorption capacity test and an absorption time test predicted the percentage of users/carers finding leakage performance acceptable, accurate to within +/- eight percentage points for all six test products. A different absorption capacity test proved most successful for bedpads used as back-up to body-worn products. It predicted the percentage of users/carers finding leakage performance acceptable, accurate to +/- five percentage points for all six products.

Improving the performance of small incontinence pads: a study of "wet comfort".

Purpose The wet comfort of small, disposable incontinence pads has been found to correlate strongly with overall acceptability. This study examined the relationship between pad properties (absorption capacity, strike‐through, and wetback) and wet comfort. Design A group of women with light incontinence were asked to report on the wet comfort of a series of experimental pads made up to the same design geometry but in different combinations of materials. Methods Three different pad variants were selected, and two experiments were conducted. Experiment 1 used a single‐blind design in which 20 testers were supplied with a random mix of the three unlabeled pad variants. All pads were saved for weighing and scored for overall performance, wet comfort, and absorbency. In experiment 2, subjects used each pad variant in turn for 1 week and at the end of each week compared that pad with the one used the previous week. Main Outcome Measures Absorption capacity, wetback, and strike‐through properties for each pad variant were measured in the laboratory setting, and wet comfort was measured by a written tool completed by participants during clinical trials. Results Data analysis from experiment 1 revealed statistically significant differences among pad variants, particularly when assessed for absorbency. Differences for wet comfort were less marked and only achieved significance when those with ratings of good or OK for wet comfort were compared with those with a score of poor. In experiment 2, no statistically significant differences were found for wet comfort. Conclusions Overall, the improvement in wet comfort achieved by a substantial increase in absorption capacity and reduction in wetback were disappointing. The relatively small sample and the design specifications of the pad may have masked differences. Cost considerations mean that more evidence is needed to justify the inclusion of more expensive materials and production stages in the manufacture of small pads.

A multicenter comparative evaluation: disposable pads for women with light incontinence

OBJECTIVE The purpose of this study was to evaluate the current range of disposable pads for women with light incontinence available in the United Kingdom. SETTING AND SUBJECTS Sixty community-based women aged 50 years or older were recruited to the study from several locations throughout the United Kingdom. METHODS A multiple crossover design was used, which enabled all subjects to test all of the 12 products that were available on the UK market at the time of the evaluation. Two tools were used to evaluate the products: a product performance questionnaire and a pad leakage diary. This incorporated both subjective and objective outcome measures. RESULTS Statistically significant differences were found between the products in 13 of the 17 performance aspects, eg, ability to hold urine without leaking and the fit of the pad. CONCLUSIONS As a group, all of the products tested performed well in terms of their ability to hold urine without leakage. However, two products performed particularly well and scored highly on most aspects of performance and represented successful “all-round” products; these were the Tena Lady Extra and the Prevail Extra Plus.

An investigation of the repeatability and reproducibility of ISO 11948-1 (the Rothwell method) for measuring the absorption capacity of incontinence pads.

The repeatability and reproducibility (precision within and between laboratories, respectively) of an international standard method (ISO 11948-1, the Rothwell method) for measuring the absorption capacity of incontinence pads was investigated. The 74 shaped disposable bodyworn insert pads for heavy incontinence on the UK market in spring 1997 were tested in three laboratories experienced in using the method, one in each of England, Spain and Sweden. Coefficients of variation (standard deviation as a proportion of the mean) for five repeats rarely exceeded 5% within any laboratory. However, there were systematic differences between laboratories: results from the Swedish and Spanish laboratories typically exceeded those from the English laboratory by 13% and 8%, respectively. The good repeatability suggests that the method is capable of adequate precision but the poor reproducibility implies that the instructions in the standard for building and/or using the test apparatus are inadequate, leaving too much room for interpretation. Having studied the data presented here and viewed videos of the apparatus in use in five laboratories (including the three contributing to this note) the ISO working group which wrote the original standard has identified several likely sources of imprecision and is now working to revise the standard to improve its reproducibility.

A multicenter evaluation of absorbent products for children with incontinence and disabilities.

OBJECTIVE To evaluate all disposable pull-up-style pads (pull-ups) designed for children with learning and or physical disabilities and compare these with a representative sample of disposable diapers (diapers). DESIGN A multicenter randomized crossover trial where all children evaluated each product. SETTING AND CHILDREN A total of 61 community dwelling children with disabilities were recruited from 5 areas throughout the United Kingdom. Families were invited by letter to participate through their local clinical nurse specialist for continence care. INSTRUMENTS Instruments used included a range of questionnaires (product performance, design performance, and design preference questionnaires) and pad weight and leakage diaries. METHODS The children tested 10 products (5 pull-ups and 5 diapers) for up to 1 week per product both at home and at school. Caregivers (parents and guardians) completed the questionnaires after evaluation of individual products and design groups had been completed. They recorded wet product weights and scored pad leakage at each pad change. School staff completed an abbreviated product performance questionnaire for each product. RESULTS There was little difference in the overall performance of the pull-ups compared with the diapers. In the pull-up group, one product performed significantly worse than the others for some performance aspects. Pull-ups worked particularly well for children able to assist with toileting and pad changes. Diapers were liked for ease of changing if the child used calipers, adapted footwear, or trousers. Diapers were more popular than pull-ups for night use, when greater absorbency was required. Neither design performed differently for school or home use. Most caregivers expressed a preference for diapers or pull-ups based on a range of individual needs, for example, level of independence in toileting, discreetness in pad change, fit, and product appearance. CONCLUSIONS Diapers and pull-ups have different strengths and limitations, and both meet the specific needs of individual children. Although pull-ups are more expensive than diapers, their use is justified if based on a thorough assessment of individual need.