P-INC: Project Incontinence- Effectiveness of Nursing Recommendations

The aim of this study is to measure the effectiveness of the implementation of evidence based nursing recommendations regarding UI in Austrian nursing homes, including three research questions:

1. What is the statistical difference between the frequency of daily UI in nursing home residents in the intervention group when compared to the control group?
2. What is the statistical difference between the frequency of documented categorisation of UI (UI diagnosis) in the intervention group (implementation of the nursing recommendations) compared to the control group (standard care)?
3. What is the statistical difference between the amounts of used nursing interventions in the intervention group compared to the control group?

5. Methods 5.1. Design A three-month cluster-randomized trial will be conducted in nursing homes in Styria and Carinthia.

5.2. Setting All nursing homes in Styria and Carinthia with a capacity ≥ 50 beds registered in a national database [15] will be invited to participate by means of an information flyer and email. With regards to practicability invitation will be started with those nursing homes with the highest number of beds.

5.3. Sample Currently there are 176 nursing homes in Styria and Carinthia with a capacity ≥ 50 beds registered in a national database with a total of 15208 nursing home residents. Through experiences from former projects a low response rate is expected. We assume a response rate of 2% on the nursing home level, amounting to 3 participating nursing homes with 259 residents. Nursing homes that want to participate voluntarily will be contacted via telephone starting at the largest for further inclusion criteria.

Inclusion criteria for nursing home residents are living in the nursing home at the date of baseline measurement and being female and assumed to stay for the whole duration of the study (3 months). Written informed consent from the participating residents or their legal representatives is required.

The nursing interventions are made for women, so we have to focus our sample to women. In Austrian nursing homes about 70% are women.Which means we have a total of 181 female nursing home residents that can be asked for participation.

We know from former studies that 90% of nursing home residents are able and willing to participate in a research study, amounting to an expected total of 163 participants in the 3 remaining nursing homes. We assume that 10% of the residents drop out during the study period. Therefore 147 female residents in the 3 nursing homes are basis for our calculation.

These 3 nursing homes will be randomly assigned (eg. via "Randomizer") to the intervention group and to the control group.

An Intra Class Coefficient (ICC) 0.05 was assumed from the literature. With the "Sample Size Calculator" developed by Campbell et al. (2004) the detectable difference with these fixed sample sizes was calculated.

With an expected cluster size of 49 nursing home residents in 3 clusters and a power of 80 % and a significance level of 0.05 we expect a minimum decrease in the frequency of daily UI from 90% to 55%.

6.0 Ethical considerations This study will be conducted in line with the declaration of Helsinki and the "Good scientific practice" of the Medical University of Graz. Nursing homes can withdraw their participation without stating reasons via telephone and email. Informed consent will be obtained of all participating residents. Residents can withdraw anytime in the study without stating reasons.

The questionnaire is based on the routine nursing care interventions and data regarding UI. Therefore no risk or side effects for the participating residents are expected. Nursing homes in the control group receive the nursing recommendations after the three month study period.

In this study only data will be collected. Therefore names of the residents cannot be traced.

7.0 Data collection including instruments Baseline measurement: The baseline measurement will be conducted to gather data on resident level (Table 1).

Table 1: Collected baseline data including time Collected data Collection time Purpose

• Participation Baseline
• Age Baseline Comparing residents
• Diagnosis Dementia Baseline Comparing residents
• Care dependency Baseline/6weeks/month Comparing residents
• Continence status Baseline/6weeks/month Comparing residents

The questionnaire at resident's level is based on the questionnaire "Pflegequalitätserhebung 2.0", which is the Austrian questionnaire of the "International Prevalence Measurement of Care Problems" research project. The Institute of Nursing Science at the Medical University of Graz is member of this research project. Questions concerning the continence status include the presence, use of catheter, staring point, frequency, diagnosis of urinary incontinence and the used nursing interventions for the management of urinary incontinence.

The German version of the Care Dependency Scale (CDS) will be used to measure the degree of care dependency with reference to different needs such as mobility or continence, where lower scores referred to higher levels of care dependency. The CDS sum score of the residents can range from completely care dependent (15-24 points), to a great extent care dependent (25-44 points), partially care dependent (45-59 points), to a limited extent care independent (60-69 points) and up to almost care independent (70-75 points).

Outcome measurement Table 2 shows the collected outcome data, the collection time and purpose. Again the Austrian version of the standardized and tested "International Prevalence Measurement of Care Problems" will be used to collect data on key outcomes, i.e. UI frequency and diagnosis.

Table 2: Collected outcome data including time and purpose Collection time Purpose Frequency of UI Baseline/6 weeks/3 month Intervention effect Diagnosis of UI Baseline/6 weeks/3 month Intervention effect Use of nursing interventions Baseline/6 weeks/3 month Intervention effect

8.0 Data analysis Data analysis via SPSS (Version 23 or higher) will be performed in consultation with the Institute for Medical Informatics, Statistics and Documentation of the Medical University of Graz, Austria, using descriptive statistics and mixed model analyses. According to Sedgwick (2015) and Shah (2011) the analysis will be conducted following "per protocol" as well as "intention to treat" analysis. The "per protocol" analysis will show the effectiveness under study conditions and the "intention to treat" will show the effect of the nursing recommendations in real life nursing practice.