Urinary incontinence in nursing home residents: a cross-national comparison - Continuing and Rehabilitative Care for Elderly People: A Comparison of Countries and Settings

Age and Ageing, Nov, 1997 by Antonio Sgadari, Eva Topinkova, Jan Bjornson, Roberto Bernabei

Introduction

Urinary incontinence (UI) is one of the most common, disruptive and often disabling conditions affecting elderly people. Between 5 and 30% of elderly people living in the community, 25-50% of hospitalized elderly inpatients and 40-50% of nursing home residents suffer from UI [1-4]. Along with its medical, psychological and social consequences, UI represents a large economic burden, increasing the costs of nursing home care [5]. In addition, nursing staff report that UI is a difficult, frustrating and demanding condition to manage [6].

A Minimum Data Set (MDS)-derived cross-national data base provides a unique opportunity (i) to ascertain whether the factors known to influence UI play the same role in different countries and (ii) to compare the different, culturally-derived UI treatment practice patterns.

The aim of the present paper is to report on the prevalence of UI in populations of elders living in continuing care settings like nursing homes in seven countries, to describe the main factors that are associated with UI and to determine how similar these and associated care practices for incontinent residents are from country to country.

Materials and methods

We analysed data from the interRAI cross-national data base of nursing home assessments, described elsewhere [7]. Data were collected from the nursing home MDS/Resident Assessment Instrument (RAI) [8] used on samples of facility residents from the following countries: Denmark, France, Iceland, Italy, Japan, Sweden and the USA. For three countries (USA, Denmark and Iceland) the data are population-based. The data sets from the remaining four countries are likely to be representative, but not as broadly-based.

A resident was considered incontinent of urine when he or she had any uncontrolled leakage of urine, regardless of amount, that occurred two or more times a week. MDS items measuring UI have been proved to be valid [9] and reliable [10]. Data on whether testing related to identifying the aetiology of UI or faecal impaction were also derived from the MDS and are defined as any of a series of tests performed in the last 30 days and noted in the record. Information on the use of selected incontinence training programmes or appliances, such as diapers, is also taken from the MDS. Cognitive status was measured by means of the cognitive performance scale (CPS), a summary score ranging from 0 (intact) to 6 (very severe impairment) based on the approach validated by Morris et al. [11]. Other independent demographic variables (age and gender) and functioning variables (mobility and being bed fast) as well as having a urinary tract infection were taken directly from the MDS. Eventually, a case-mix index (CMI) summary measure was used using the RUG-III classification system, the CMI is a relative measure of the daily resource use spent on nursing home residents, based on their characteristics. The higher the CMI for a given country, the heavier case-mix of population residing in nursing homes in that country [12]. Statistical analysis was performed using SAS statistical software. The association between UI and residents' characteristics was analysed by means of the [chi square]-test separately for each country; a P value lower than 0.05 was considered significant.

Results

The present analysis was performed on seven national samples for a total sample of 279 191 elderly residents. Table 1 shows the main characteristics of the population samples used. The age distribution is very skewed toward older ages in all the samples, with more than about 40% of residents being 85 years old or older in all the countries but Japan (35%). Women represent a remarkable proportion of the total population (65.9-76.2%), with somewhat higher figures in Italy (85.7%). Physical disability is consistent across countries, with a mean ADL index ranging from 8.2 in Japan and Denmark to 10.5 in France. High percentages of residents show signs of mild to very severe mental deterioration as measured by the CPS. Finally, the CMI is lower in Denmark (0.72), higher in Sweden (1.02) and ranges between 0.76 and 0.90 in the other samples.

Table 1. Characteristics of the population samples

                                           Country

                              Denmark   France   Iceland   Italy

Age (years, %)
 [is less than] 75            15.1      25.2     11.2      18.2
             75-84            34.8      30.5     37.3      42.4
              85              50.1      44.3     51.5      39.3
Women (%)                     76.2      69.9     67.1      85.7
Mean ADL index                 8.2      10.5      9.2       9.7
Cognitively impaired (%)(a)   59.7      72.0     54.4      66.5
Mean case-mix index            0.72      0.87     0.76      0.87

                                      Country

                              Japan   Sweden   USA

Age (years, %)
 [is less than] 75            18.6    18.3     18.9
             75-84            46.3    38.1     33.9
             85               35.2    43.5     47.2
Women (%)                     71.7    65.9     73.2
Mean ADL index                 8.2    --        9.9
Cognitively impaired (%)(a)   53.5    67.3     65.0
Mean case-mix index            0.83    1.02    0.90