Final Report Summary - SHELTER (Services and Health for Elderly in Long TERm care)
Executive Summary:
One of the greatest challenges to comparing residents and evaluating nursing home care between countries is that valid standardized methods for recording care needs are not in routine daily use in Europe. The Minimum Data Set (MDS) is a comprehensive care needs assessment instrument containing a set of standardized essential clinical and functional status measures that enable comparison of resident characteristics, measures of quality of care and resident outcomes. It provides the data for a system of outcome management and quality assurance for elders receiving long term care. We proposed to conduct a study that would validate the MDS assessment instrument for long term care residents (MDS-LTCF) as a methodology for comparing the provision of care in NH in Europe.
To ensure that the assessment instrument is relevant for the wider European population, this study 1. assessed linguistic validity of the instrument in different translations of items included in the translated versions of the instrument, 2. evaluated test-retest and inter-rater reliability of each item, 3. implemented the MDS LTCF on a large scale, leading to the creation of a database that allows to measure residents outcomes, identify predictors of outcomes, develop eligibility criteria and resource utilization groupings, monitor service delivery and analyse indicators of quality of care.
1. Linguistic validity
The original version of the MDS LTCF was translated from English version into 7 languages of participating countries (Czech, Dutch, Finnish, French, German, Hebrew, Italian) following several steps: forward translation, reconciliation, back translation, back translation review, cognitive debriefing, review of cognitive debriefing results and proof-reading. Following this step, in each country, a team of nurses, native speakers of the target languages, were asked to review items included in the translated versions of the MDS LTCF.
2. Reliability
Overall, reliability of 201 MDS LTCF items were tested. Test-retest reliability was evaluated in 380 residents by 48 assessors in 17 facilities. Test retest reliability of MDS LTCF items in each of the assessment areas ranged from 0.75 to 0.92. Average weighted kappa for the 198 categorical items was 0.83. None of the single items in the MDS LTCF presented with a kappa value below 0.40.
Inter-rater reliability was evaluated in 404 residents by 76 assessors in 23 facilities. Inter-rater reliability of items in the assessment areas ranged from 0.64 to 0.91. Average weighted kappa for the 198 categorical items was 0.74.
3. Implementation of MDS LTCF
A total number of 4156 residents was enrolled in the study, 500 from Czech Republic, 507 from England, 484 from Finland, 493 from France, 496 from Germany, 580 from Israel, 548 from Italy and 548 from the Netherlands. Mean age was above 80 years and women represented approximately 3/4 of the sample. Disability was common, with more than 80% of participating residents requiring assistance or being dependent in ADL and cognitive impairment was present in more than 2/3 of the sample with 30% of residents being classified as severely impaired. Elevated level of “clinical” complexity of residents was confirmed by the high prevalence of urinary incontinence (73.5%), pain (36.0%), depression (32.0%), behavioral symptoms (27.5%), falls (18.6%) and pressure ulcers (10.4%).
In conclusion, the SHELTER study validates the MDS LTCF as a methodology that can be used not only to assist clinicians in performing a comprehensive assessment of NH patients, but also to compare characteristics of older adults across countries and to collect information and create databases that can be used to answer relevant questions in the field of long term care. This new methodology can be applied on a large scale to analyze the provision of long-term care across European health systems.
Project Context and Objectives:
The SHELTER study was aimed to validate the MDS assessment instrument for long term care residents (MDS-LTCF) as a methodology to assess the provision of care in NH in Europe. This novel instrument has been developed building on the previously available MDS-like instrument but with a European perspective. To make this instrument applicable to a large European population, this study assessed a. linguistic validity of the instrument in different translations; b. face validity of items included in the translated versions of the instrument; c. test-retest and inter-rater reliability of each item. In addition, the SHELTER study proposes to implement and apply on al large scale the MDS LTCF, leading to d. the creation of a database
The Project plan was based on the following steps:
1. At month 1 (January 2009) a kick off meeting was held in Rome. During this meeting, project activities were defined and organized;
2. The first activity concerned the linguistic validation of the assessment form from the original English version into 7 target languages (Czech, Dutch, Finnish, French, German, Hebrew, Italian). This process followed several steps: forward translation, reconciliation, back translation, back translation review, cognitive debriefing, review of cognitive debriefing results and proof-reading. This procedure is in line with international recommendations. Following this step, in each country, a team of nurses, native speakers of the target languages, were asked to review items included in the translated versions of the MDS LTCF and to judge whether they reflect the content they are intended to have. The team was composed by eight nurses in most countries, with the only exception of Germany where only five nurses were in the team;
3. The European Nursing Home Services assessment tool (EU-NHS) was prepared. This tool includes data on included data on structural and organizational characteristics of the facility, including number of beds and staff characteristics, characteristics of the area, including the demographic characteristics of geographic area, number of hospital and/or residential care beds per thousand people, financial characteristics, including private/public, for profit or non-profit, sources of funding, total budget and system of budget allocation and co-payment, management structures, range and organisation of the services provided, including interaction with primary and secondary health care services and team meetings, presence of dedicated special care units, including dementia unit, hospice, dialysis, service delivery including eligibility criteria, referral systems, provision of integrated service, health/social professionals and administrative personnel per patient, total number of residents per year, duration of service provision per patient, waiting lists availability, use of assessment instruments and application of guidelines, comprehensive geriatric assessment, presence of case managers, team meetings, care planning and use of multidisciplinary team approach;
4. An Intranet network for data sharing and data input will be set up and tested shortly after the kick off meeting (see website www.shelter-elderly.eu);
5. At month 7 a pre-enrolment meeting was held in Paris (July 2009) and study partners agreed on study protocol and standardized the training procedure;
6. After this meeting each partner finalized the preliminary agreements with all the nursing homes considered for the study and performed the training of study personnel. Assessors responsible for data collection were trained following a previously validated procedure. In each country, 2-day courses were organised to educate assessors about the concepts of comprehensive geriatric assessment and multidisciplinary teamwork and to train them to the use of MDS LTCF. In the first part of the course main gerontological and geriatric problems were presented through classes, seminars and conferences. The concept of comprehensive geriatric assessment and multidisciplinary teamwork were introduced and multidimensional assessment tools were presented. In the second part of the course participants learned how to perform the assessment using the MDS LTCF, including the specific forms and appropriate response codes, and to develop care planning. Assessors were trained to use a variety of information sources, such as direct observation, interviews with the person under care, family, friends, or formal service providers, and review clinical records, both medical and nursing. The assessors were ordinary clinical staff, external research staff, or a mixture of both. Most assessors were nurses, but other professionals were also used. In line with interRAI's standard approach to coding, they were all instructed to exercise their best clinical judgment in order to record observations based on their evaluation of the most accurate information source;
7. At month 10 the baseline assessments of participating residents began and the EU-NHS forms were completed. In each country, study partners identified a sample of NH willing to participate to the study. This sample must not be considered randomly selected and is not intended to be representative of all NH in each country. The number of participating facilities in each country is presented in figure 1 (see attached documents).
4156 older adults were assessed using the MDS LTCF. Participants enrolled in the study were re-assessed at 6 and 12 months if still in the facility. If no longer in the facility, reason (death, hospitalization, discharge at home or in another institution) and date of death or discharge were recorded;
8. Among study participants test-retest reliability was evaluated in 380 residents by 48 assessors in 17 facilities. Inter-rater reliability was evaluated in 404 residents by 76 assessors in 23 facilities;
9. A mid-term meeting was held at month 14 in Treviso (February 2010). In this meeting, the partners reviewed the progress of the project, identify possible problems and define solutions and programmed the dissemination activities;
10. Data regarding resident’s characteristics and resource consumption were electronically stored using a WEB/WAP based application. The EU-NHS data will be collected using dedicated software developed by the co-ordinating centre and distributed to each partner. The resulting database (Shelter database) was used to perform analyses to measure residents outcomes, identify predictors of outcomes, develop eligibility criteria and resource utilization groupings, monitor service delivery and analyse indicators of quality of care.
11. The final meeting was held in Brussels (December 2011) and attended by Project partners and by national policy makers, health administrators and politicians responsible for health care decisions.
Results of reliability analyses and characteristics of residents enrolled in the study
Overall, reliability of 201 InterRAI LTCF items was tested. Only 3 items (walking speed, height and weight) were continuous variables. Test-retest reliability was evaluated in 380 residents by 48 assessors (nurses: 42; MD: 5, psychologist: 1) in 17 facilities. This sample was composed of 52 dual assessments from Italy, 50 from Czech Republic, Germany and Israel, 48 from England and France, 43 from Finland and 39 from the Netherlands. As shown in table 2, test retest reliability of InterRAI LTCF items in each of the assessment areas ranged from 0.75 to 0.92. Average weighted kappa for the 198 categorical items was 0.83. None of the single items in the InterRAI LTCF presented with a kappa value below 0.40. Pearson’s correlation coefficients for items based on continuous variables were 0.96 for walking speed, 0.92 for height and 0.96 for weight.
Inter-rater reliability was evaluated in 404 residents by 76 assessors (nurses: 68; MD: 4; psychologists: 4) in 23 facilities. This sample was composed of 60 dual assessments from Israel, 59 from Italy, 50 from England, 49 from France and Czech Republic, 48 from Germany and Finland and 41 from the Netherlands. As shown in Table 2 inter-rater reliability of items in the assessment areas ranged from 0.64 to 0.91. Average weighted kappa for the 198 categorical items was 0.74. Only one single item had a kappa value below 0.40 (‘Fluid output exceeds input’ included in the Oral/Nutritional Status assessment area). Pearson’s correlation coefficients for items based on continuous variables were 0.88 for walking speed, 0.99 for height and 0.89 for weight.
A total number of 4156 residents was enrolled in the study, 500 from Czech Republic, 507 from England, 484 from Finland, 493 from France, 496 from Germany, 580 from Israel, 548 from Italy and 548 from the Netherlands. Mean age was above 80 years and women represented approximately 3/4 of the sample. Disability was common, with more than 80% of participating residents requiring assistance or being dependent in ADL and cognitive impairment was present in more than 2/3 of the sample with 30% of residents being classified as severely impaired. Elevated level of “clinical” complexity of residents was confirmed by the high prevalence of urinary incontinence (73.5%), pain (36.0%), depression (32.0%), behavioral symptoms (27.5%), falls (18.6%) and pressure ulcers (10.4%).
Dissemination activities
The dissemination of projects results was tackled in different ways:
1. Organization in symposia during scientific meetings including the VII EUROPEAN INTERNATIONAL CONGRESS of the INTERNATIONAL ASSOCIATION OF GERONTOLOGY AND GERIATRICS (IAGG) - HEALTHY AND ACTIVE AGEING FOR ALL EUROPEANS held in Bologna (Italy), April 14-17, 2011, the 2008 meeting of the Italian Society of Gerontology and Geriatrics (SIGG), the 2010 ECHE meeting in Helsinki (FI);
2. Submission of manuscripts to scientific peer reviewed journals;
3. The final conference of the project entitled ‘Long Term Care of Europe’s Older Citizens: Policy and Fact. Providing data to support the relationship between care practice organization and quality of care in long term care facilities in Europe’ that was held in Brussels on December 5, 2011 constituted “per se” a form of dissemination activity. It was attended by the project partners and by stakeholders, national policy makers, health administrators and politicians responsible for health decisions. This meeting constituted a way to make many opinion leaders in the health services for the elderly field aware on relevant areas of policy related to SHELTER findings and illustrate the policy problems for providing high quality care for frail older people in NH.
4. A one page promotional summary of the project was produced to announce its aims and generate interest among a broad audience. A more detailed overview was provided in a project identity brochure, drawing on background information from the SHELTER project.
5. A website (www.shelter-elderly.eu) was established and regularly updated with the newest information about the project. The SHELTER website describes the aims of the project and collected all relevant documents related to the project, including the final conference program, the promotional summary the brochure and the slides presented in the final meeting.
Project Results:
Workpackage 0.1. Project management
WP leader: partner 1 - UNIVERSITA’ CATTOLICA SACRO CUORE (UCSC)
Objective: To perform administrative and financial tasks towards the Commission and among project partners.
Status of the WP: completed
The SHELTER management tasks scheduled to be performed during the period 01/01/2009 (M1)-30/06/2010 (M18) were successfully conducted and completed in accordance with the Annex I–Description of Work dated October 2007.
All the WP0.1 (Project management) Achievement have all been successfully achieved and delivered.
Indeed:
- The final Consortium Agreement was signed, and sent out to all Beneficiaries and all financial issues, including cost statement, preparation and submission were managed at the Coordinating Center (Task 0.1);
- The kick off Meeting was organized and held in Rome (IT) on January 15th and 16th, 2009 (Task 0.2);
- A telephone hot-line was established at the Co-ordinating centre (Task 0.3);
- The pre-enrolment Meeting was organized and held in Paris (FR) on July 4th and 5th, 2009 (Task 0.4);
- The mid-term Meeting was organized and held in Treviso (IT) on February 25th and 26th, 2010 (Task 0.5);
- The Final Consortium Meeting was organized and held in Brussels (B) on December 5th, 2011 (Task 0.6).
Workpackage 0.2. Scientific coordination and quality assurance
WP leader: partner 1 - UNIVERSITA’ CATTOLICA SACRO CUORE (UCSC)
Objective: To establish a project quality plan and to ensure optimal development of partner co-operation and outcome of project efforts and to co-ordinate the project partners and the tasks to be carried out.
Status of the WP: completed
All the 0.2 (Scientific coordination and quality assurance) Achievement due in the reporting period M1-M18 have all been successfully achieved and delivered.
- Members of the Quality Management Committee were nominated during the kick –off meeting (Task 0.7). The Quality Management (QM) Committee is responsible for: a. monitor and review the quality of Project Achievements; b. confirm that Project Milestones are in fact achieved and reached on time; c. confirm that WP and activity objectives are being addressed and met by the WPs; d. confirm that any revision to the Project Plan is feasible and in accordance with directives from the SC; d. help to ensure adherence to the decision procedures, reporting routines and communication flows as defined within the Project Organisation. As decided during the kick-off meeting the QMC is represented by principal scientists of each study site (Prof. Bernabei, Prof. Topinkova, Prof. Frijters, Prof. Carpenter, Prof Henrard, Prof Gindin, Prof. Nikolaus and Prof. Finne-Soveri). The Project Manager is responsible for administrative and financial issues. During the kick-off meeting Dr. Mariotti was nominated Project manager. The interaction among these structures is summarised in figure 2 (see attached documents).
- Periodic Conference calls of the Steering Committee were held (Task 0.8);
- An Intranet connection for data exchange and Shelter website (see WP 3) were developed (Task 0.9). A website was established and regularly updated at the following address: http://www.shelter-elderly.eu. The website includes:
1. a public area which includes a description of the rational, design and methodology of the Shelter project, together with a list of the partners involved in the project. Through the public area of the website it is possible to access the Shelter software and to collect EU-NHS form (by the use of a specific log-in and password). Main findings of the Shelter project will also be published in the public area of the website;
2. a private area, accessible by the partners through the use of log-in and passwords provided by the coordinating center. In the private area documents and presentations relevant to the project are made available, as well as an agenda relative to the Shelter project and project’s news. In addition through this area it is possible to submit to the coordinating center proposal for secondary analyses involving the use of the Shelter database.
- Regular visits to the participating sites (see WP 7) by a member of the coordinating centre were organized (Task 0.10).
Workpackage 1. Linguistic validation
WP leader: partner 4 - NATIONAL INSTITUTE FOR HEALTH AND WELFARE (THL)
Objective: To translate the English version MDS-LTCF into official languages of the participating countries
Status of the WP: completed
Aim of the WP1 was to translate the MDS-LTCF from the original English version into 7 target languages (Italian, French, Finnish, Hebrew, Czech, German and Dutch). Before starting the translation process, during the kick-off meeting, partners discussed the need to bring changes to country specific items included in the form. In particular, the following items of the MDS-LTCF form were modified in order to make them applicable to a nursing home residents in Europe, irrespective of country of residence: Section A, items 5 and 7, Section B, items 3 and 4, Section P, items 1 and 2.
Once these changes were made, the translation process was started based on a multistep approach:
Step 1. Forward translation. Two forward translations into each target language were carried out by two independent translators, native speakers of the target language and fluent in the source language.
Step 2. Reconciliation. The reconciliation of the two forward translations was carried out by an independent translator, native speaker of the target language, together with the responsible investigator for each country. At this step, discrepancies between the original independent translations were resolved.
Step 3. Back translation. Back translation was carried out by one professional translator, native speaker of the source language and fluent in the target language. Back translators did not have prior knowledge of the MDS form.
Step 4. Back translation review. The responsible investigator for each country reviewed back translations against the source language, identified possible discrepancies and eventually required further revised versions of reconciled translations.
Step 5. Cognitive debriefing. A cognitive debriefing was performed by the responsible investigator separately for each country to assess the level of comprehensibility and cognitive equivalence of the translation. Eight nurses for each country, native speakers of the target languages were included in the cognitive debriefing.
Step 6. Review of cognitive debriefing results. A review of the cognitive debriefing results was performed by the responsible investigator for each country in order to incorporate findings of the debriefing process.
Step 7. Proof-reading. The responsible investigator for each country carried out the proof-reading of the final translations as the final quality-control step.
Then, face validity was assessed as a qualitative measure of validity. After linguistic validation, a sample of eight nurses native speakers of the target languages in each country was asked to read items included in the translated versions of the MDS and make their judgements on whether those items reflect the content they are intended to have.
This task (T1.1) was completed in the expected time (month 4) and translated versions of the MDS-LTCF form were made available for implementation.
Achievements
Versions of the MDS-LTCF translated in official languages of countries involved in the study (7 files) are available for examination by the commission (A1).
Workpackage 2. Design of the EU-NHS form
WP leader: partner 9 - VERENIGING VOOR CHRISTELIJK HOGER ONDERWIJS WETENSCHAPPELIJK ONDERZOEK EN PATIENTENZORG (VUA)
Objective: to design the European Nursing Home Services assessment form (EU-NHS).
Status of the WP: completed
The EU-NHS for was designed and developed to evaluate and assess the characteristics of the Nursing Home Services/Agencies participating to the project (Task 2.1). A first version of the EUNHS form was designed by the coordinating center before the kick off meeting. This form included data on:
- structural and organizational characteristics of the facility, including number of beds and staff characteristics ;
- characteristics of the area, including the demographic characteristics of geographic area, number of hospital and/or residential care beds per thousand people;
- financial characteristics, including private/public, for profit or non-profit, sources of funding, total budget and system of budget allocation and co-payment;
- management structures, range and organisation of the services provided, including interaction with primary and secondary health care services and team meetings;
- presence of dedicated special care units, including dementia unit, hospice, dialysis, etc.;
- service delivery including eligibility criteria, referral systems, provision of integrated service, health/social professionals and administrative personnel per patient, total number of residents per year, duration of service provision per patient, waiting lists availability, use of assessment instruments and application of guidelines, comprehensive geriatric assessment, presence of case managers, team meetings, care planning and use of multidisciplinary team approach.
This first version was widely discussed during the kick off meeting (month 1) and changes were made to the form in order to make it applicable to nursing homes in all countries participating to the study and to make possible a comprehensive assessment of characteristics of nursing homes. Six versions of the form followed the first one. Each version was discussed by participating partners either by conference calls or e-mail. Agreement was obtained on the eighth version of the form (Task 2.2) and this was discussed again and revised for minor editing and approved by all partners during the pre-enrolment meeting (Month 7, Task 2.3).
The delay in the elaboration of the final version of the form (expected to be completed at month 4, completed at month 7), did not determine a delay in the overall development of the project, because training of researchers to the use of the MDS LTCF was planned and performed at month 8.
Achievements
Final version of The European Nursing Home Services (EU-NHS) assessment form and the former versions of the form (8 files) are available for examination by the commission (A2).
Workpackage 3. Intranet network implementation
WP leader: partner 1 - UNIVERSITA’ CATTOLICA SACRO CUORE (UCSC)
Objective: to set up and test an Intranet network for data sharing among partners
Status of the WP: completed
A subcontractors was responsible for the creation of an Intranet Network, for data entry and data sharing among participants and the creation of a website. The selection of the subcontractor was carried out according to the Università Cattolica del Sacro Cuore usual practice. The principle of best price-quality ratio, under conditions of transparency and equal treatment, was respected. Criteria for subcontracting indicated by 7FP guidelines were fulfilled.
Information needed were identified by the partners and the subcontractor during the kick off meeting (Task 3.1). The Intranet Network consists of a central server that can be accessed via secure links over Internet by light clients used by the participants (Task 3.2). Light clients can either be a PC with a standard internet WEB browser, a smart-phone with a WAP browser (Task 3.3). The connection can be established using an ADSL link or its wireless equivalent, a GSM/GPRS connection.
A special software was developed exclusively for the Shelter project (Task 3.4). This software allows:
1. To collect the MDS-LTCF data for each resident enrolled in the study at different time points (baseline, 6 months and 12 months). The data collected through the software populates the Shelter database;
2. To obtain a personal health profile (PHP) for each resident enrolled in this study. The PHP is obtained through information included in the MDS-LTCF and it can be used for clinical purposes. The following areas are assessed in the PHP: Mental Health, Communication, Behaviour Patterns, Social Functioning, Physical Functioning, Medical Complexity, Pain, Skin Care, Nutritional status, Special Treatments;
3. To identify problematic areas of the resident through the Clinical Assessment Protocols (CAPs), a standardized procedure that assists the assessor to interpret systematically all the information recorded on an instrument and help the clinician focus on key issues identified during the assessment process, so that decisions as to whether and how to intervene can be explored with the individual;
4. To extract data raw collected into the Shelter database in an aggregated way to perform analyses.
Both PHP and CAP obtained through the software are available for clinical use for study researchers. Specific instructions on the use of the software were released during the pre-enrolment meeting and available for study researchers. Partners and study researchers received personal log-in and password to access the software and tested it (Task 3.6). Security was implemented by means of SSL and WTLS technologies. Process of data entry and database implementation is described in figure 3 (see attached documents).
In addition, a website was established and regularly updated at the following address: Task 3.5 and 3.7. The website includes:
1. a public area which includes a description of the rational, design and methodology of the Shelter project, together with a list of the partners involved in the project. Through the public area of the website it is possible to access the Shelter software and to collect EU-NHS form (by the use of a specific log-in and password). Main findings of the Shelter project will also be published in the public area of the website;
2. a private area, accessible by the partners through the use of log-in and passwords provided by the coordinating center. In the private area documents and presentations relevant to the project are made available, as well as an agenda relative to the Shelter project and project’s news. In addition through this area it is possible to submit to the coordinating center proposal for secondary analyses involving the use of the Shelter database
Achievements
Files with report on the Project Intranet Architecture (4 files), instructions for the use of the software, examples of CAP and PHP (A3) are available for examination by the commission.
Workpackage 4. Project protocol and Standard training definition
WP leader: partner 2 - CHARLES UNIVERSITY – FIRST MEDICAL FACULTY (CUNI)
Objective: to organise the Project pre-enrolment meeting and to plan and define a standardised training for the assessors who will collect data on residents (MDS LTCF) and on NH Services characteristics (EU-NHS).
Status of the WP: completed
During the pre-enrolment meeting the project protocol was defined and standardization of training procedure was discussed. In addition partners were introduced to the EU-NHS form (Task 4.1). The discussion of the project protocol was focused on the following key issues:
ENROLMENT AND FU TARGET POPULATION
Overall 500 participants will be enrolled in each site. Based on the project protocol residents should be re-assessed at 6 and 12 months. Given the necessity to reassess patients an effort is required to study partners to oversample long stay residents.
RELIABILITY DATA
Inter-rater reliability: InterRAI LTCF will be administered to 50 residents by two independent raters in each country (possibly in the same day).
Test-retest reliability: InterRAI LTCF will be administered to 50 residents on two different occasions (possibly the 2 assessment should be performed within 48 hours, 72 hours at the most).
It is key the health conditions of the resident are not changed between the 2 assessment.
DISCHARGE ASSESSMENT
Residents assessment will be performed at 6 and 12 months. An additional assessment will be filled-in when the participant is discharged from the facility, hospitalized or dead. In case of death it will not be necessary to fill in the whole assessment, but only a part of it, indicating the date of death
As already mentioned also standardization of training was discussed and standardized (Task 4.2). Nationally, courses were be organised to:
• Give nurses theoretical bases in the main gerontological and geriatric problems, along with instructions on the concept of comprehensive geriatric assessment and multidisciplinary teamwork;
• Train nurses to the use of the MDS LTCF;
In the first part of the course main gerontological and geriatric problems should be presented through classes, seminars and conferences. The concept of comprehensive geriatric assessment and multidisciplinary teamwork will be introduced and multidimensional assessment tools will be presented.
In the second part of the course participants should learn how to perform MDS assessment, including the specific forms and appropriate coding. They must also develop a comprehensive understanding of the connection between MDS and the Care Planning Process. Strategies for effective care planning must be discussed. Teachers will carefully explain:
1. The intent of the listed in the MDS LTCF, explaining criteria for inclusion, discussions of how the information will be used by clinical staff to identify person problems and develop care plan.
2. Supplemental definitions, examples, and instructions for completing MDS LTCF items.
3. Reminders of which MDS LTCF items require observation of the person for other than the standard 3-day observation period.
4. Sources of information to be consulted in completing specific MDS LTCF items.
5. Coding, explaining proper method of recording each response, with explanations of individual response categories. Coding of the following items will be analyse in depth: Identification Information, Demographic information, Cognitive patterns, Communication/hearing patterns, Vision patterns, Abuse and neglect, Mood and behaviour patterns, Psychosocial well-being, Physical functioning and structural problems, Continence, Disease diagnoses, Health conditions, Oral/nutritional status, Oral/dental status, Skin condition, Activity pursuit patterns, Medications, Special treatments and procedures, Discharge potential and overall status, Assessment information, Medications, Resident assessment protocol summary.
A simulated case must be presented and discussed in group.
Audiovisuals may be used including transparencies, slides and videotapes to present simulated Services and Health for Elderly in Long TERm care cases. At the end of the courses a final evaluation of participating nurses will be performed.
Participating nurses should also be invited to evaluate teachers and write their feedbacks on specifically designed anonymous forms.
Additionally a training course must be organized to train research assistants to the use of the EU-NHS form. During this course the EU-NHS must be explained and practical example will be provided.
MDS LTCF manual (English version) was made available for research nurses training and translated by participated partners if necessary (see file MDS LTCF manual).
Achievements
Files on Shelter Project final protocol, guidelines on training, pre-enrolment meeting
Minutes, MDS LTCF manual (A4) and one page summary of the project, posters for a non specialist audience (A5) are available for examination by the commission.
Workpackage 5. Training
WP leader: partner 1 – UNIVERSITA’ CATTOLICA DEL SACRO CUORE (UCSC)
Objective: to train research nurses to perform the assessment and data collection.
Status of the WP: completed
Courses to train research nurses to collect data on residents enrolled in the study were organised locally (Task 5.1). Training was performed by participating partners locally following guidelines discussed during the pre-enrolment meeting. The training was based on a 5-days, 5 hour a day course. The course consists of two parts, the first one on the basic concepts of gerontology and geriatrics and on the theories of geriatric nursing; the second is a practical training where the participants learn the use of the Minimum Data Set. Both sections are based on the MDS – LTCF. This course has been tested before, and the results obtained have been published on international journals (The MDS instrument as teaching tool for nursing home staff, Educat Geront 1996; Variation in training programs for Resident Assessment Instrument implementation, Age Ageing 1997; A brief training program on RAI instrument improves motivation of nursing home staff, Fact Research Gerontol 1998). The details of the classes are as follows:
1st day: Introduction to the course, principal theories on aging, demographic and epidemiological aspects of aging, multi-dimensional assessment and principal multi-dimensional assessment tools;
2nd day: Introduction to the Minimum Data Set – Long Term care Facility, analysis of the items of the sections AA through H, practical training on the evaluation of the Activities of the Daily Living;
3rd day: Analysis of the items of the sections H through Q, practical training on the whole assessment form, introduction to the Client Assessment Protocols;
4th day: Use of the Minimum Data Set to identify and evaluate the principal geriatric syndromes: delirium, cognitive impairment/dementia, communication/hearing, vision, physical functioning / activities of daily living, instrumental activities of daily living;
5th day: Use of the Minimum Data Set to identify and evaluate the principal geriatric syndromes: rehabilitation potential, bladder continence / indwelling catheter, psychosocial well-being, mood status.
At the end of the course, the participants were able to perform a multi-dimensional evaluation of the client using the Minimum Data Set. They were also be able to design an individual care plan allowing to maximize and/or maintain the best possible level of physical functioning and psychosocial well-being. Simulated cases were used to evaluate information retained from the course (Task 5.2).
Locally research assistant were trained to the use of the EUNHS form in a 1-day course (Task 5.3). For each participating facility a research assistant responsible for collecting data of the EUNHS form was identified.
Achievements
Internal report on procedures for nurse training (IR1) is available for examination by the commission.
Workpackage 6. Residents enrolment and follow up
WP leader: partner 8 – UNIVERSITY OF KENT AT CANTERBURY (UniKent)
Objective: to standardise the procedures and to start the enrolment phase.
Status of the WP: completed
Since the MDS-LTCF instrument is supposed to be used to assess nursing home residents irrespective of age, diseases and health condition, no specific exclusion criterium was used to select residents participating to the study. Only inclusion criterium was being a nursing home resident. In addition, partners were suggested to oversample, if possible, long stay residents because a 6 and 12 months follow up was planned for the project.
Residents were enrolled from different nursing homes in participating countries. Every nursing home in the project must be registered in the website and EUNHS form completed.
Finally, since the aim of the SHELTER project is to validate a methodology to analyze the provision of care in nursing home in Europe and not to compare characteristics of NH residents and systems it was decided not to randomize facilities and patients in order to have a country representative sample.
Resident enrolment has started at month 10 (October 2009). Overall, 4156 residents from participating countries (original target was 4000 residents) were enrolled and completed the baseline assessment (Task 6.1 and 6.2). In addition, 784 residents were enrolled to calculate test-retest reliability and inter-rater reliability. Test-retest reliability was evaluated in 380 residents by 48 assessors (nurses: 42; MD: 5, psychologist: 1) in 17 facilities (Task 6.3). This sample was composed of 52 dual assessments from Italy, 50 from Czech Republic, Germany and Israel, 48 from England and France, 43 from Finland and 39 from the Netherlands. Inter-rater reliability was evaluated in 404 residents by 76 assessors (nurses: 68; MD: 4; psychologists: 4) in 23 facilities (Task 6.4). This sample was composed of 60 dual assessments from Israel, 59 from Italy, 50 from England, 49 from France and Czech Republic, 48 from Germany and Finland and 41 from the Netherlands. The baseline assessment was completed at month 13 (December 2009). Six and twelve months follow up assessments were completed by December 2010 (Task 6.6 and 6.7). In addition, residents leaving the facility before the end of the study were reassessed by study staff (discharge assessments). Number of residents completed the baseline, follow up and discharge assessments by country are listed in the table 1 (see attached documents).
During the follow up data on mortality, including date, place and cause of death are collected by study nurses as well as data on hospital admissions and length of stay, emergency room visits, hospital outpatient referral, and day care (Task 6.8).
Participating residents were enrolled in 57 facilities. For each facility the EUNHS form was completed and data collected through the study website (Task 6.5). Geographic distribution of participating facilities is presented in figure 1 (see attached documents).
Achievements
Procedures and criteria for patient enrolment (1 file - IR2) are available for examination by the commission.
Workpackage 7. Study monitoring
WP leader: partner 1 – UNIVERSITA’ CATTOLICA DEL SACRO CUORE (UCSC)
Objective: to ensure both standardisation and timely progress through all the study stages.
Status of the WP: completed
The appointed partner’s 1 (UCSC) staff member, with considerable experience in this field having already co-ordinated at a European level studies on implementation of comprehensive assessment instruments, visited the participating centres in order to review the progress in data collection, as well as to detect, discuss and solve with local partners any problem arising regarding residents and Services assessment, data interpretation, collection or storage (Task 7.1 and 7.2). Quality assurance was also ensured during site visits with the production of site Gantt chart to schedule all the foreseen activities. Partner’s 1 staff visited each study site twice every year during the phase of enrolment and follow up. All information from such visits were analysed by the Steering Committee in accordance with the procedures detailed in the workplan and in the management section (Task 7.3). Monitoring of the activities of participating facilities was performed locally by project partners. Reports on data collection progress and problem assessment were discussed during the mid term meeting and Steering Committee meetings. Tables reporting status of the enrolment procedures in different countries were provided to partners and constantly updated. The following problems were identified during the mid-term meeting:
Partner 4 (THL) did not provide the translated version of the InterRAI LTCF in doc format. No data on assessments were available in the software for the following partners: Partners 4 (THL), 5 (AP-HP.SPR) and 9 (VUA).
The partners were committed to complete their duties as follows:
Partner 9 (VUA): baseline and test-retes and inter-rater reliability data made available by end of March, 2010
Partner 5 (AP-HP.SPR): baseline and test-retes and inter-rater reliability data made available by end of May, 2010
Partner 4 (THL): translation of InterRAI LTCF and baseline and test-retes and inter-rater reliability data made available by end of April, 2010. Now these problems are fully solved and the project does not present any delay.
Achievements
Gantt charts to monitor study progress (IR3), Mid-term meeting report indicating problems on data collection progress and assessment (A6) and Steering Committee reports indicating problems on data collection progress and assessment (A7 and A8) are available for examination by the commission.
Workpackage 8. Shelter database implementation
WP leader: partner 9 – VERENIGING VOOR CHRISTELIJK HOGER ONDERWIJS WETENSCHAPPELIJK ONDERZOEK EN PATIENTENZORG (VUA)
Objective: to create the Shelter database, trough the collection of the following information: a. residents baseline and follow-up (6 and 12 months) data collected by the MDS LTCF; b. data on reliability; c. structural and organizational characteristics of the nursing home Services, collected by the EU-NHS form; d. data on mortality and health care resource consumption.
Status of the WP: completed
Residents-level data were collected at baseline, six and twelve months by research nurses trained on the use of MDS LTCF in each country. In addition, to evaluate reliability of the instrument dual, independent assessments will be performed in a sample of 100 residents in each country. Study researchers responsible for data entry were provided with password to access database and data entry (Task 8.1).
A variety of validation checks were applied during the data entry process (Task 8.2). These included intra-form and inter-form checks for consistency of responses, range checking, and checks for required fields and skip patterns. More sophisticated checks were made after the data have been entered. Errors and inconsistencies detected at this stage may generate queries for the clinic staff.
Data entry took taking place via an internet-based web browser interface (Task 8.3). Electronic data were managed centrally at the Co-ordinating Centre, with a feedback mechanism through e-mail and web-based reporting. Raw files were generated from this process (Task 8.4).
Achievements
First set of analytical files collecting data on baseline assessment, drug use and facility characteristics (A9), a second (A10), third (A11) and final set (A12) of analytical files are available for examination by the commission.
Workpackage 9. Data analysis
WP leader: partner 8 - UNIVERSITY OF KENT AND CANTERBURY (UniKent)
Objective: To analyse the reliability of items included in the MDS LTCF; the baseline demographic, clinical and functional characteristics of the Nursing Home residents in the selected countries; quality of care provided in the NH included in the study; resident’s outcomes, to identify predictors (both at a patient level and at a service level) of mortality, health resources consumption and other functional and clinical outcomes.
Status of the WP: completed
Analytical files were created for data analysis (Task 9.1) and the following analyses were performed:
Reliability analysis (Task 9.2)
Test-retest reliability and Inter-rater reliability of items included in the MDS LTCF were measured estimating Kappa statistics for nominal data (Kappa values below 0.4 were considered indicators of poor reliability, values between 0.4 and 0.75 were considered indicators of adequate reliability and values of 0.75 or above were considered indicators of excellent reliability). Pearson’s correlation coefficient (r) were calculated to measure agreement for continuous and ordinal data respectively.
Overall, reliability of 201 InterRAI LTCF items was tested. Only 3 items (walking speed, height and weight) were continuous variables. Table 2 (see attached documents) presents average weighted kappa values for categorical items in different assessment areas. Test-retest reliability was evaluated in 380 residents by 48 assessors (nurses: 42; MD: 5, psychologist: 1) in 17 facilities. This sample was composed of 52 dual assessments from Italy, 50 from Czech Republic, Germany and Israel, 48 from England and France, 43 from Finland and 39 from the Netherlands. As shown in table below, test retest reliability of InterRAI LTCF items in each of the assessment areas ranged from 0.75 to 0.92. Average weighted kappa for the 198 categorical items was 0.83. None of the single items in the InterRAI LTCF presented with a kappa value below 0.40. Pearson’s correlation coefficients for items based on continuous variables were 0.96 for walking speed, 0.92 for height and 0.96 for weight.
Inter-rater reliability was evaluated in 404 residents by 76 assessors (nurses: 68; MD: 4; psychologists: 4) in 23 facilities. This sample was composed of 60 dual assessments from Israel, 59 from Italy, 50 from England, 49 from France and Czech Republic, 48 from Germany and Finland and 41 from the Netherlands. As shown in table below inter-rater reliability of items in the assessment areas ranged from 0.64 to 0.91. Average weighted kappa for the 198 categorical items was 0.74. Only one single item had a kappa value below 0.40 (‘Fluid output exceeds input’ included in the Oral/Nutritional Status assessment area). Pearson’s correlation coefficients for items based on continuous variables were 0.88 for walking speed, 0.99 for height and 0.89 for weight.
Description of baseline characteristics of study sample (Task 9.3)
As mentioned, total number of 4156 residents was enrolled in the study, 500 from Czech Republic, 507 from England, 484 from Finland, 493 from France, 496 from Germany, 580 from Israel, 548 from Italy and 548 from the Netherlands. Table 3 (see attached documents) summarizes the principal characteristics of NH residents enrolled in the study. Mean age was above 80 years and women represented approximately 3/4 of the sample. Disability was common, with more than 80% of participating residents requiring assistance or being dependent in ADL and cognitive impairment was present in more than 2/3 of the sample with 30% of residents being classified as severely impaired. Elevated level of “clinical” complexity of residents was confirmed by the high prevalence of urinary incontinence (73.5%), pain (36.0%), depression (32.0%), behavioral symptoms (27.5%), falls (18.6%) and pressure ulcers (10.4%).
Analysis of quality of care (Task 9.4)
The MDS LTCF produces Quality Indicators (QIs) relevant for long term care. Some of these QI are summarized in table 4 (see attached documents).
In particular, some QI were analysed in depth. For example, an ‘adhoc’ analysis was dedicated to polypharmacy (concurrent use of 5-9 drugs) and excessive polypharmacy (concurrent use of ≥ 10 drugs) .
Polypharmacy was observed in 2000 (49.7%) residents and excessive polypharmacy in 979 (24.3%) residents. As shown in Figure 4 (see attached documents), prevalence of polypharmacy and excessive polypharmacy widely varied in study sample. In particular, Italy had the lowest prevalence of excessive polypharmacy (8.8%), followed by Israel (12.9%), Germany (15.7%), England (22.7%), the Netherlands (24.4%), Czech Republic (25.2%), France (30.2%) and Finland (56.7%).
As shown in table 5 (see attached documents) residents on polypharmacy and excessive polypharmacy, when compared with those not on polypharmacy, had a less severe level of disability and cognitive impairment, a reduced rate of behavioural symptoms and a higher prevalence of depression, falls, pain, dyspnoea, GI symptoms, dizziness and flare up of a recurrent or chronic problem. Finally, residents on polypharmacy and excessive polypharmacy presented a higher number of concomitant diseases and, in particular, ischemic heart disease, heart failure, Parkinson disease, stroke, diabetes and cancer were more common in these groups as compared with the non-polypharmacy group.
Table below reports results of the multivariate analysis, identifying variables independently associated with polypharmacy and excessive polypharmacy. As compared with non-polypharmacy, excessive polypharmacy was directly associated with depression (OR 1.81; 95% CI 1.38-2.37) pain, dyspnoea, GI symptoms, and specific diseases including ischemic heart disease, heart failure, Parkinson disease, stroke and diabetes, while an inverse association was shown for age, ADL disability and cognitive impairment. Similarly, polypharmacy was directly associated with depression, pain, ischemic heart disease, heart failure, Parkinson disease and diabetes. Presence of behavioural symptoms presented an inverse and significant association with polypharmacy, but not with excessive polypharmacy.
Facilities characteristics - Resources utilization (Task 9.5)
Characteristics of participating facilities were analysed using data from the EUNHS form. These data are summarized in table 6.
MDS LTCF assessment systems support the use of algorithms that group individuals into categories reflecting the relative costs of services they are likely to consume. These tools are designed to address an enormous challenge faced by governments the world over: how to allocate limited resources to persons in need in a fair and equitable way. Case mix is by definition a system that classifies people into groups that are homogeneous in their use of resources. A good case-mix system also gives meaningful clinical descriptions of these individuals. The application of case mix is broad; it provides the basis, not only for reimbursement, but also for comparing facilities or programs, practice patterns, as an adjunct to quality of care and efficiency measurement, a staff planning tool, etc.
The Resource Utilization Groups (RUG-III) case-mix algorithm was developed to provide a patient-specific means of allocating health care resources based on the variable costs of caring for individuals with different needs. The RUG-III uses 108 variables from the MDS LTCF to create 44 34 categories of patients with homogeneous resource use patterns. The RUG-III algorithm explains about 55% of variance in resource use, and it has been validated in a number of countries through a series of international studies. Figures 5 and 6 (see attached documents) present the RUG-III distribution of residents in the SHELTER study.
Identify independent predictors of residents outcomes (Task 9.6)
Participating residents were followed for 1 year. Of the 4156 residents in the baseline samples, follow up data were collected for 3741 residents (drop out rate 10%). Overall 761 residents died during the follow up (20.3%). Cox analysis was perform to identify independent predictors of mortality. Results of this analysis is presented in the table 7 (see attached documents).
Workpackage 10. Dissemination of project results
WP leader: partner 3 - UNIVERSITY OF ULM (UUlm)
Objective: to disseminate the Project’s results
Status of the WP: completed
The dissemination of projects results was tackled in different ways:
1. Organization in symposia during scientific meetings (Task 10.1):
a. A symposium was organized during the VII EUROPEAN INTERNATIONAL CONGRESS of the INTERNATIONAL ASSOCIATION OF GERONTOLOGY AND GERIATRICS (IAGG) - HEALTHY AND ACTIVE AGEING FOR ALL EUROPEANS held in Bologna (Italy), April 14-17, 2011. The Symposium entitled ‘LONG-TERM CARE CHALLENGES IN EUROPE - STANDARDIZED ASSESSMENT FOR ADDRESSING PATIENT’S NEEDS, OUTCOMES AND QUALITY OF CARE’ included four talks focused on results of the SHELTER study;
b. During the 2008 meeting of the Italian Society of Gerontology and Geriatrics (SIGG) a special symposium on ‘EUROPEAN RESEARCH PROJECTS ON HEALTH SYSTEMS AND LONG TERM CARE OF THE ELDERLY’ was held. The symposium involved representatives from the Shelter project as well as investigators from other projects funded under the FP7, the Interlinks project and the Ancien project;
c. During the 2010 ECHE meeting in Helsinki (FI), during a session on ‘Long-Term Care for the Elderly: A European Perspective’ the design, methodology a status of the Shelter project were presented.
2. Submission of manuscripts to scientific peer reviewed journals (Task 10.2). So far the following manuscripts based on SHELTER data were published:
a. Onder G, Carpenter I, Finne-Soveri H, Gindin J, Frijters D, Henrard JC, Nikolaus T, Topinkova E, Tosato M, Liperoti R, Landi F, Bernabei R. Assessment of nursing home residents in Europe: the Services and Health for Elderly in Long TERm care (SHELTER) study. BMC Health Serv Res. 2012 Jan 9;12(1):5.
b. Onder G, Liperoti R, Fialova D, Topinkova E, Tosato M, Danese P, Gallo PF, Carpenter I, Finne-Soveri H, Gindin J, Bernabei R, Landi F; for the SHELTER Project. Polypharmacy in Nursing Home in Europe: Results From the SHELTER Study. J Gerontol A Biol Sci Med Sci. 2012 Jan 4.
c. Tosato M, Lukas A, van der Roest HG, Danese P, Antocicco M, Finne-Soveri H, Nikolaus T, Landi F, Bernabei R, Onder G. Association of pain with behavioral and psychiatric symptoms among nursing home residents with cognitive impairment: Results from the SHELTER study. Pain. 2012 Feb;153(2):305-10
3. The final conference of the project entitled ‘Long Term Care of Europe’s Older Citizens: Policy and Fact. Providing data to support the relationship between care practice organization and quality of care in long term care facilities in Europe’ that was held in Brussels on December 5, 2011 constituted “per se” a form of dissemination activity. It was attended by the project partners and by stakeholders, national policy makers, health administrators and politicians responsible for health decisions. This meeting constituted a way to make many opinion leaders in the health services for the elderly field aware on relevant areas of policy related to SHELTER findings and illustrate the policy problems for providing high quality care for frail older people in NH.
4. A one page promotional summary of the project was produced to announce its aims and generate interest among a broad audience. A more detailed overview was provided in a project identity brochure, drawing on background information from the SHELTER project.
5. A website (www.shelter-elderly.eu) was established and regularly updated with the newest information about the project. The SHELTER website describes the aims of the project and collected all relevant documents related to the project, including the final conference program, the promotional summary the brochure and the slides presented in the final meeting.
Achievements
Program of the symposium held at the meeting of the INTERNATIONAL ASSOCIATION OF GERONTOLOGY AND GERIATRICS (IAGG – A18), as well as programs of the meeting of the Italian Society of Gerontology and Geriatrics and of the ECHE meeting, papers accepted for publication (A19), the program of the final conference, the brochure and the flyer are available for examination by the commission.
Potential Impact:
Impact
The present project validates a new methodology for analysing provision of long term care in Europe. To date, no comprehensive assessment tool specifically targeting European long term care industry has been developed and validated. As such, this is the first attempt to demonstrate the efficacy of a scientific methodology for understanding current services and care processes in European nursing homes, addressing the needs of individual residents and providing evidence for resource allocation, planning and organization.
The ageing of the population is leading health care expenditure to be one of the greatest socioeconomic issues for European countries. The ageing phenomenon will be responsible for a quite dramatic societal change with less working age individuals available to take care of an instead progressively larger number of disabled and dependent elderly individuals. Under this scenario, the long term care industry is projected to rapidly expand with a substantial impact on public spending and on planning of health care services by the governments of the States Members.
This foreseeable public health emergency requires a strategic plan using a scientific approach.
The main barrier for an efficient public health intervention in this field is represented by the nearly total lack of knowledge about long-term care services across Europe. Information about structural and organizational characteristics of long term care services is rather sparse, and documentation extremely heterogeneous and influenced by cultural and societal expectations. In those countries where the system is regulated, extremely different and not comparable standards are adopted.
This applies to eligibility criteria, assessment instruments, protocols and procedures, financing mechanisms. Thus, the implementation of a standardized, comprehensive, cross-languages and cross–cultures assessment instrument is believed to be a crucial, initial step toward the acquisition of the necessary knowledge base.
The current project validates the MDS tool as the first comprehensive assessment instrument, specifically intended for the application within the European nursing home system. Scientific evidence was produced to nursing home operators, medical directors, administrators, policy makers, public health experts, health care professionals and researchers that the MDS represents a valid and readily applicable instrument. A large scale implementation across Europe could now be promoted. The systematic collection of information with the MDS instrument led to the creation of a cross-national database including data on nursing home residents and services characteristics. A systematic collection of data using the MDS tool provided reliable and detailed cross-national data on nursing home industry in Europe. For the first time, countries share a common language to understand, analyse and manage what is going to be the most critical sector of the health care industry in the near future.
Data sharing gets European countries to learn from each other and to manage diversity and heterogeneity in respect of a unique high standard of quality and cost-effectiveness. Such methodology produces convincing evidence concerning clinical outcomes of individuals or groups of residents, eligibility criteria, to ensure monitoring of services delivery, to measure quality and to estimate efficacy and effectiveness of prevention programs. Also, costeffectiveness of different health care strategies from different countries were scientifically compared and the best model of care may be identified. An ultimate impact on the reduction of public spending for long term care and on the rationalization of resources allocation in health care is much expected.
Dissemination
The dissemination of projects results was tackled in different ways:
1. Organization in symposia during scientific meetings:
a. A symposium was organized during the VII EUROPEAN INTERNATIONAL CONGRESS of the INTERNATIONAL ASSOCIATION OF GERONTOLOGY AND GERIATRICS (IAGG) - HEALTHY AND ACTIVE AGEING FOR ALL EUROPEANS held in Bologna (Italy), April 14-17, 2011. The Symposium entitled ‘LONG-TERM CARE CHALLENGES IN EUROPE - STANDARDIZED ASSESSMENT FOR ADDRESSING PATIENT’S NEEDS, OUTCOMES AND QUALITY OF CARE’ included four talks focused on results of the SHELTER study;
b. During the 2008 meeting of the Italian Society of Gerontology and Geriatrics (SIGG) a special symposium on ‘EUROPEAN RESEARCH PROJECTS ON HEALTH SYSTEMS AND LONG TERM CARE OF THE ELDERLY’ was held. The symposium involved representatives from the Shelter project as well as investigators from other projects funded under the FP7, the Interlinks project and the Ancien project;
c. During the 2010 ECHE meeting in Helsinki (FI), during a session on ‘Long-Term Care for the Elderly: A European Perspective’ the design, methodology a status of the Shelter project were presented.
2. Submission of manuscripts to scientific peer reviewed journals. So far the following manuscripts based on SHELTER data were published:
a. Onder G, Carpenter I, Finne-Soveri H, Gindin J, Frijters D, Henrard JC, Nikolaus T, Topinkova E, Tosato M, Liperoti R, Landi F, Bernabei R. Assessment of nursing home residents in Europe: the Services and Health for Elderly in Long TERm care (SHELTER) study. BMC Health Serv Res. 2012 Jan 9;12(1):5.
b. Onder G, Liperoti R, Fialova D, Topinkova E, Tosato M, Danese P, Gallo PF, Carpenter I, Finne-Soveri H, Gindin J, Bernabei R, Landi F; for the SHELTER Project. Polypharmacy in Nursing Home in Europe: Results From the SHELTER Study. J Gerontol A Biol Sci Med Sci. 2012 Jan 4.
c. Tosato M, Lukas A, van der Roest HG, Danese P, Antocicco M, Finne-Soveri H, Nikolaus T, Landi F, Bernabei R, Onder G. Association of pain with behavioral and psychiatric symptoms among nursing home residents with cognitive impairment: Results from the SHELTER study. Pain. 2012 Feb;153(2):305-10
3. The final conference of the project entitled ‘Long Term Care of Europe’s Older Citizens: Policy and Fact. Providing data to support the relationship between care practice organization and quality of care in long term care facilities in Europe’ that was held in Brussels on December 5, 2011 constituted “per se” a form of dissemination activity. It was attended by the project partners and by stakeholders, national policy makers, health administrators and politicians responsible for health decisions. This meeting constituted a way to make many opinion leaders in the health services for the elderly field aware on relevant areas of policy related to SHELTER findings and illustrate the policy problems for providing high quality care for frail older people in NH.
4. A one page promotional summary of the project was produced to announce its aims and generate interest among a broad audience. A more detailed overview was provided in a project identity brochure, drawing on background information from the SHELTER project.
5. A website (www.shelter-elderly.eu) was established and regularly updated with the newest information about the project. The SHELTER website describes the aims of the project and collected all relevant documents related to the project, including the final conference program, the promotional summary the brochure and the slides presented in the final meeting.
One of the greatest challenges to comparing residents and evaluating nursing home care between countries is that valid standardized methods for recording care needs are not in routine daily use in Europe. The Minimum Data Set (MDS) is a comprehensive care needs assessment instrument containing a set of standardized essential clinical and functional status measures that enable comparison of resident characteristics, measures of quality of care and resident outcomes. It provides the data for a system of outcome management and quality assurance for elders receiving long term care. We proposed to conduct a study that would validate the MDS assessment instrument for long term care residents (MDS-LTCF) as a methodology for comparing the provision of care in NH in Europe.
To ensure that the assessment instrument is relevant for the wider European population, this study 1. assessed linguistic validity of the instrument in different translations of items included in the translated versions of the instrument, 2. evaluated test-retest and inter-rater reliability of each item, 3. implemented the MDS LTCF on a large scale, leading to the creation of a database that allows to measure residents outcomes, identify predictors of outcomes, develop eligibility criteria and resource utilization groupings, monitor service delivery and analyse indicators of quality of care.
1. Linguistic validity
The original version of the MDS LTCF was translated from English version into 7 languages of participating countries (Czech, Dutch, Finnish, French, German, Hebrew, Italian) following several steps: forward translation, reconciliation, back translation, back translation review, cognitive debriefing, review of cognitive debriefing results and proof-reading. Following this step, in each country, a team of nurses, native speakers of the target languages, were asked to review items included in the translated versions of the MDS LTCF.
2. Reliability
Overall, reliability of 201 MDS LTCF items were tested. Test-retest reliability was evaluated in 380 residents by 48 assessors in 17 facilities. Test retest reliability of MDS LTCF items in each of the assessment areas ranged from 0.75 to 0.92. Average weighted kappa for the 198 categorical items was 0.83. None of the single items in the MDS LTCF presented with a kappa value below 0.40.
Inter-rater reliability was evaluated in 404 residents by 76 assessors in 23 facilities. Inter-rater reliability of items in the assessment areas ranged from 0.64 to 0.91. Average weighted kappa for the 198 categorical items was 0.74.
3. Implementation of MDS LTCF
A total number of 4156 residents was enrolled in the study, 500 from Czech Republic, 507 from England, 484 from Finland, 493 from France, 496 from Germany, 580 from Israel, 548 from Italy and 548 from the Netherlands. Mean age was above 80 years and women represented approximately 3/4 of the sample. Disability was common, with more than 80% of participating residents requiring assistance or being dependent in ADL and cognitive impairment was present in more than 2/3 of the sample with 30% of residents being classified as severely impaired. Elevated level of “clinical” complexity of residents was confirmed by the high prevalence of urinary incontinence (73.5%), pain (36.0%), depression (32.0%), behavioral symptoms (27.5%), falls (18.6%) and pressure ulcers (10.4%).
In conclusion, the SHELTER study validates the MDS LTCF as a methodology that can be used not only to assist clinicians in performing a comprehensive assessment of NH patients, but also to compare characteristics of older adults across countries and to collect information and create databases that can be used to answer relevant questions in the field of long term care. This new methodology can be applied on a large scale to analyze the provision of long-term care across European health systems.
Project Context and Objectives:
The SHELTER study was aimed to validate the MDS assessment instrument for long term care residents (MDS-LTCF) as a methodology to assess the provision of care in NH in Europe. This novel instrument has been developed building on the previously available MDS-like instrument but with a European perspective. To make this instrument applicable to a large European population, this study assessed a. linguistic validity of the instrument in different translations; b. face validity of items included in the translated versions of the instrument; c. test-retest and inter-rater reliability of each item. In addition, the SHELTER study proposes to implement and apply on al large scale the MDS LTCF, leading to d. the creation of a database
The Project plan was based on the following steps:
1. At month 1 (January 2009) a kick off meeting was held in Rome. During this meeting, project activities were defined and organized;
2. The first activity concerned the linguistic validation of the assessment form from the original English version into 7 target languages (Czech, Dutch, Finnish, French, German, Hebrew, Italian). This process followed several steps: forward translation, reconciliation, back translation, back translation review, cognitive debriefing, review of cognitive debriefing results and proof-reading. This procedure is in line with international recommendations. Following this step, in each country, a team of nurses, native speakers of the target languages, were asked to review items included in the translated versions of the MDS LTCF and to judge whether they reflect the content they are intended to have. The team was composed by eight nurses in most countries, with the only exception of Germany where only five nurses were in the team;
3. The European Nursing Home Services assessment tool (EU-NHS) was prepared. This tool includes data on included data on structural and organizational characteristics of the facility, including number of beds and staff characteristics, characteristics of the area, including the demographic characteristics of geographic area, number of hospital and/or residential care beds per thousand people, financial characteristics, including private/public, for profit or non-profit, sources of funding, total budget and system of budget allocation and co-payment, management structures, range and organisation of the services provided, including interaction with primary and secondary health care services and team meetings, presence of dedicated special care units, including dementia unit, hospice, dialysis, service delivery including eligibility criteria, referral systems, provision of integrated service, health/social professionals and administrative personnel per patient, total number of residents per year, duration of service provision per patient, waiting lists availability, use of assessment instruments and application of guidelines, comprehensive geriatric assessment, presence of case managers, team meetings, care planning and use of multidisciplinary team approach;
4. An Intranet network for data sharing and data input will be set up and tested shortly after the kick off meeting (see website www.shelter-elderly.eu);
5. At month 7 a pre-enrolment meeting was held in Paris (July 2009) and study partners agreed on study protocol and standardized the training procedure;
6. After this meeting each partner finalized the preliminary agreements with all the nursing homes considered for the study and performed the training of study personnel. Assessors responsible for data collection were trained following a previously validated procedure. In each country, 2-day courses were organised to educate assessors about the concepts of comprehensive geriatric assessment and multidisciplinary teamwork and to train them to the use of MDS LTCF. In the first part of the course main gerontological and geriatric problems were presented through classes, seminars and conferences. The concept of comprehensive geriatric assessment and multidisciplinary teamwork were introduced and multidimensional assessment tools were presented. In the second part of the course participants learned how to perform the assessment using the MDS LTCF, including the specific forms and appropriate response codes, and to develop care planning. Assessors were trained to use a variety of information sources, such as direct observation, interviews with the person under care, family, friends, or formal service providers, and review clinical records, both medical and nursing. The assessors were ordinary clinical staff, external research staff, or a mixture of both. Most assessors were nurses, but other professionals were also used. In line with interRAI's standard approach to coding, they were all instructed to exercise their best clinical judgment in order to record observations based on their evaluation of the most accurate information source;
7. At month 10 the baseline assessments of participating residents began and the EU-NHS forms were completed. In each country, study partners identified a sample of NH willing to participate to the study. This sample must not be considered randomly selected and is not intended to be representative of all NH in each country. The number of participating facilities in each country is presented in figure 1 (see attached documents).
4156 older adults were assessed using the MDS LTCF. Participants enrolled in the study were re-assessed at 6 and 12 months if still in the facility. If no longer in the facility, reason (death, hospitalization, discharge at home or in another institution) and date of death or discharge were recorded;
8. Among study participants test-retest reliability was evaluated in 380 residents by 48 assessors in 17 facilities. Inter-rater reliability was evaluated in 404 residents by 76 assessors in 23 facilities;
9. A mid-term meeting was held at month 14 in Treviso (February 2010). In this meeting, the partners reviewed the progress of the project, identify possible problems and define solutions and programmed the dissemination activities;
10. Data regarding resident’s characteristics and resource consumption were electronically stored using a WEB/WAP based application. The EU-NHS data will be collected using dedicated software developed by the co-ordinating centre and distributed to each partner. The resulting database (Shelter database) was used to perform analyses to measure residents outcomes, identify predictors of outcomes, develop eligibility criteria and resource utilization groupings, monitor service delivery and analyse indicators of quality of care.
11. The final meeting was held in Brussels (December 2011) and attended by Project partners and by national policy makers, health administrators and politicians responsible for health care decisions.
Results of reliability analyses and characteristics of residents enrolled in the study
Overall, reliability of 201 InterRAI LTCF items was tested. Only 3 items (walking speed, height and weight) were continuous variables. Test-retest reliability was evaluated in 380 residents by 48 assessors (nurses: 42; MD: 5, psychologist: 1) in 17 facilities. This sample was composed of 52 dual assessments from Italy, 50 from Czech Republic, Germany and Israel, 48 from England and France, 43 from Finland and 39 from the Netherlands. As shown in table 2, test retest reliability of InterRAI LTCF items in each of the assessment areas ranged from 0.75 to 0.92. Average weighted kappa for the 198 categorical items was 0.83. None of the single items in the InterRAI LTCF presented with a kappa value below 0.40. Pearson’s correlation coefficients for items based on continuous variables were 0.96 for walking speed, 0.92 for height and 0.96 for weight.
Inter-rater reliability was evaluated in 404 residents by 76 assessors (nurses: 68; MD: 4; psychologists: 4) in 23 facilities. This sample was composed of 60 dual assessments from Israel, 59 from Italy, 50 from England, 49 from France and Czech Republic, 48 from Germany and Finland and 41 from the Netherlands. As shown in Table 2 inter-rater reliability of items in the assessment areas ranged from 0.64 to 0.91. Average weighted kappa for the 198 categorical items was 0.74. Only one single item had a kappa value below 0.40 (‘Fluid output exceeds input’ included in the Oral/Nutritional Status assessment area). Pearson’s correlation coefficients for items based on continuous variables were 0.88 for walking speed, 0.99 for height and 0.89 for weight.
A total number of 4156 residents was enrolled in the study, 500 from Czech Republic, 507 from England, 484 from Finland, 493 from France, 496 from Germany, 580 from Israel, 548 from Italy and 548 from the Netherlands. Mean age was above 80 years and women represented approximately 3/4 of the sample. Disability was common, with more than 80% of participating residents requiring assistance or being dependent in ADL and cognitive impairment was present in more than 2/3 of the sample with 30% of residents being classified as severely impaired. Elevated level of “clinical” complexity of residents was confirmed by the high prevalence of urinary incontinence (73.5%), pain (36.0%), depression (32.0%), behavioral symptoms (27.5%), falls (18.6%) and pressure ulcers (10.4%).
Dissemination activities
The dissemination of projects results was tackled in different ways:
1. Organization in symposia during scientific meetings including the VII EUROPEAN INTERNATIONAL CONGRESS of the INTERNATIONAL ASSOCIATION OF GERONTOLOGY AND GERIATRICS (IAGG) - HEALTHY AND ACTIVE AGEING FOR ALL EUROPEANS held in Bologna (Italy), April 14-17, 2011, the 2008 meeting of the Italian Society of Gerontology and Geriatrics (SIGG), the 2010 ECHE meeting in Helsinki (FI);
2. Submission of manuscripts to scientific peer reviewed journals;
3. The final conference of the project entitled ‘Long Term Care of Europe’s Older Citizens: Policy and Fact. Providing data to support the relationship between care practice organization and quality of care in long term care facilities in Europe’ that was held in Brussels on December 5, 2011 constituted “per se” a form of dissemination activity. It was attended by the project partners and by stakeholders, national policy makers, health administrators and politicians responsible for health decisions. This meeting constituted a way to make many opinion leaders in the health services for the elderly field aware on relevant areas of policy related to SHELTER findings and illustrate the policy problems for providing high quality care for frail older people in NH.
4. A one page promotional summary of the project was produced to announce its aims and generate interest among a broad audience. A more detailed overview was provided in a project identity brochure, drawing on background information from the SHELTER project.
5. A website (www.shelter-elderly.eu) was established and regularly updated with the newest information about the project. The SHELTER website describes the aims of the project and collected all relevant documents related to the project, including the final conference program, the promotional summary the brochure and the slides presented in the final meeting.
Project Results:
Workpackage 0.1. Project management
WP leader: partner 1 - UNIVERSITA’ CATTOLICA SACRO CUORE (UCSC)
Objective: To perform administrative and financial tasks towards the Commission and among project partners.
Status of the WP: completed
The SHELTER management tasks scheduled to be performed during the period 01/01/2009 (M1)-30/06/2010 (M18) were successfully conducted and completed in accordance with the Annex I–Description of Work dated October 2007.
All the WP0.1 (Project management) Achievement have all been successfully achieved and delivered.
Indeed:
- The final Consortium Agreement was signed, and sent out to all Beneficiaries and all financial issues, including cost statement, preparation and submission were managed at the Coordinating Center (Task 0.1);
- The kick off Meeting was organized and held in Rome (IT) on January 15th and 16th, 2009 (Task 0.2);
- A telephone hot-line was established at the Co-ordinating centre (Task 0.3);
- The pre-enrolment Meeting was organized and held in Paris (FR) on July 4th and 5th, 2009 (Task 0.4);
- The mid-term Meeting was organized and held in Treviso (IT) on February 25th and 26th, 2010 (Task 0.5);
- The Final Consortium Meeting was organized and held in Brussels (B) on December 5th, 2011 (Task 0.6).
Workpackage 0.2. Scientific coordination and quality assurance
WP leader: partner 1 - UNIVERSITA’ CATTOLICA SACRO CUORE (UCSC)
Objective: To establish a project quality plan and to ensure optimal development of partner co-operation and outcome of project efforts and to co-ordinate the project partners and the tasks to be carried out.
Status of the WP: completed
All the 0.2 (Scientific coordination and quality assurance) Achievement due in the reporting period M1-M18 have all been successfully achieved and delivered.
- Members of the Quality Management Committee were nominated during the kick –off meeting (Task 0.7). The Quality Management (QM) Committee is responsible for: a. monitor and review the quality of Project Achievements; b. confirm that Project Milestones are in fact achieved and reached on time; c. confirm that WP and activity objectives are being addressed and met by the WPs; d. confirm that any revision to the Project Plan is feasible and in accordance with directives from the SC; d. help to ensure adherence to the decision procedures, reporting routines and communication flows as defined within the Project Organisation. As decided during the kick-off meeting the QMC is represented by principal scientists of each study site (Prof. Bernabei, Prof. Topinkova, Prof. Frijters, Prof. Carpenter, Prof Henrard, Prof Gindin, Prof. Nikolaus and Prof. Finne-Soveri). The Project Manager is responsible for administrative and financial issues. During the kick-off meeting Dr. Mariotti was nominated Project manager. The interaction among these structures is summarised in figure 2 (see attached documents).
- Periodic Conference calls of the Steering Committee were held (Task 0.8);
- An Intranet connection for data exchange and Shelter website (see WP 3) were developed (Task 0.9). A website was established and regularly updated at the following address: http://www.shelter-elderly.eu. The website includes:
1. a public area which includes a description of the rational, design and methodology of the Shelter project, together with a list of the partners involved in the project. Through the public area of the website it is possible to access the Shelter software and to collect EU-NHS form (by the use of a specific log-in and password). Main findings of the Shelter project will also be published in the public area of the website;
2. a private area, accessible by the partners through the use of log-in and passwords provided by the coordinating center. In the private area documents and presentations relevant to the project are made available, as well as an agenda relative to the Shelter project and project’s news. In addition through this area it is possible to submit to the coordinating center proposal for secondary analyses involving the use of the Shelter database.
- Regular visits to the participating sites (see WP 7) by a member of the coordinating centre were organized (Task 0.10).
Workpackage 1. Linguistic validation
WP leader: partner 4 - NATIONAL INSTITUTE FOR HEALTH AND WELFARE (THL)
Objective: To translate the English version MDS-LTCF into official languages of the participating countries
Status of the WP: completed
Aim of the WP1 was to translate the MDS-LTCF from the original English version into 7 target languages (Italian, French, Finnish, Hebrew, Czech, German and Dutch). Before starting the translation process, during the kick-off meeting, partners discussed the need to bring changes to country specific items included in the form. In particular, the following items of the MDS-LTCF form were modified in order to make them applicable to a nursing home residents in Europe, irrespective of country of residence: Section A, items 5 and 7, Section B, items 3 and 4, Section P, items 1 and 2.
Once these changes were made, the translation process was started based on a multistep approach:
Step 1. Forward translation. Two forward translations into each target language were carried out by two independent translators, native speakers of the target language and fluent in the source language.
Step 2. Reconciliation. The reconciliation of the two forward translations was carried out by an independent translator, native speaker of the target language, together with the responsible investigator for each country. At this step, discrepancies between the original independent translations were resolved.
Step 3. Back translation. Back translation was carried out by one professional translator, native speaker of the source language and fluent in the target language. Back translators did not have prior knowledge of the MDS form.
Step 4. Back translation review. The responsible investigator for each country reviewed back translations against the source language, identified possible discrepancies and eventually required further revised versions of reconciled translations.
Step 5. Cognitive debriefing. A cognitive debriefing was performed by the responsible investigator separately for each country to assess the level of comprehensibility and cognitive equivalence of the translation. Eight nurses for each country, native speakers of the target languages were included in the cognitive debriefing.
Step 6. Review of cognitive debriefing results. A review of the cognitive debriefing results was performed by the responsible investigator for each country in order to incorporate findings of the debriefing process.
Step 7. Proof-reading. The responsible investigator for each country carried out the proof-reading of the final translations as the final quality-control step.
Then, face validity was assessed as a qualitative measure of validity. After linguistic validation, a sample of eight nurses native speakers of the target languages in each country was asked to read items included in the translated versions of the MDS and make their judgements on whether those items reflect the content they are intended to have.
This task (T1.1) was completed in the expected time (month 4) and translated versions of the MDS-LTCF form were made available for implementation.
Achievements
Versions of the MDS-LTCF translated in official languages of countries involved in the study (7 files) are available for examination by the commission (A1).
Workpackage 2. Design of the EU-NHS form
WP leader: partner 9 - VERENIGING VOOR CHRISTELIJK HOGER ONDERWIJS WETENSCHAPPELIJK ONDERZOEK EN PATIENTENZORG (VUA)
Objective: to design the European Nursing Home Services assessment form (EU-NHS).
Status of the WP: completed
The EU-NHS for was designed and developed to evaluate and assess the characteristics of the Nursing Home Services/Agencies participating to the project (Task 2.1). A first version of the EUNHS form was designed by the coordinating center before the kick off meeting. This form included data on:
- structural and organizational characteristics of the facility, including number of beds and staff characteristics ;
- characteristics of the area, including the demographic characteristics of geographic area, number of hospital and/or residential care beds per thousand people;
- financial characteristics, including private/public, for profit or non-profit, sources of funding, total budget and system of budget allocation and co-payment;
- management structures, range and organisation of the services provided, including interaction with primary and secondary health care services and team meetings;
- presence of dedicated special care units, including dementia unit, hospice, dialysis, etc.;
- service delivery including eligibility criteria, referral systems, provision of integrated service, health/social professionals and administrative personnel per patient, total number of residents per year, duration of service provision per patient, waiting lists availability, use of assessment instruments and application of guidelines, comprehensive geriatric assessment, presence of case managers, team meetings, care planning and use of multidisciplinary team approach.
This first version was widely discussed during the kick off meeting (month 1) and changes were made to the form in order to make it applicable to nursing homes in all countries participating to the study and to make possible a comprehensive assessment of characteristics of nursing homes. Six versions of the form followed the first one. Each version was discussed by participating partners either by conference calls or e-mail. Agreement was obtained on the eighth version of the form (Task 2.2) and this was discussed again and revised for minor editing and approved by all partners during the pre-enrolment meeting (Month 7, Task 2.3).
The delay in the elaboration of the final version of the form (expected to be completed at month 4, completed at month 7), did not determine a delay in the overall development of the project, because training of researchers to the use of the MDS LTCF was planned and performed at month 8.
Achievements
Final version of The European Nursing Home Services (EU-NHS) assessment form and the former versions of the form (8 files) are available for examination by the commission (A2).
Workpackage 3. Intranet network implementation
WP leader: partner 1 - UNIVERSITA’ CATTOLICA SACRO CUORE (UCSC)
Objective: to set up and test an Intranet network for data sharing among partners
Status of the WP: completed
A subcontractors was responsible for the creation of an Intranet Network, for data entry and data sharing among participants and the creation of a website. The selection of the subcontractor was carried out according to the Università Cattolica del Sacro Cuore usual practice. The principle of best price-quality ratio, under conditions of transparency and equal treatment, was respected. Criteria for subcontracting indicated by 7FP guidelines were fulfilled.
Information needed were identified by the partners and the subcontractor during the kick off meeting (Task 3.1). The Intranet Network consists of a central server that can be accessed via secure links over Internet by light clients used by the participants (Task 3.2). Light clients can either be a PC with a standard internet WEB browser, a smart-phone with a WAP browser (Task 3.3). The connection can be established using an ADSL link or its wireless equivalent, a GSM/GPRS connection.
A special software was developed exclusively for the Shelter project (Task 3.4). This software allows:
1. To collect the MDS-LTCF data for each resident enrolled in the study at different time points (baseline, 6 months and 12 months). The data collected through the software populates the Shelter database;
2. To obtain a personal health profile (PHP) for each resident enrolled in this study. The PHP is obtained through information included in the MDS-LTCF and it can be used for clinical purposes. The following areas are assessed in the PHP: Mental Health, Communication, Behaviour Patterns, Social Functioning, Physical Functioning, Medical Complexity, Pain, Skin Care, Nutritional status, Special Treatments;
3. To identify problematic areas of the resident through the Clinical Assessment Protocols (CAPs), a standardized procedure that assists the assessor to interpret systematically all the information recorded on an instrument and help the clinician focus on key issues identified during the assessment process, so that decisions as to whether and how to intervene can be explored with the individual;
4. To extract data raw collected into the Shelter database in an aggregated way to perform analyses.
Both PHP and CAP obtained through the software are available for clinical use for study researchers. Specific instructions on the use of the software were released during the pre-enrolment meeting and available for study researchers. Partners and study researchers received personal log-in and password to access the software and tested it (Task 3.6). Security was implemented by means of SSL and WTLS technologies. Process of data entry and database implementation is described in figure 3 (see attached documents).
In addition, a website was established and regularly updated at the following address: Task 3.5 and 3.7. The website includes:
1. a public area which includes a description of the rational, design and methodology of the Shelter project, together with a list of the partners involved in the project. Through the public area of the website it is possible to access the Shelter software and to collect EU-NHS form (by the use of a specific log-in and password). Main findings of the Shelter project will also be published in the public area of the website;
2. a private area, accessible by the partners through the use of log-in and passwords provided by the coordinating center. In the private area documents and presentations relevant to the project are made available, as well as an agenda relative to the Shelter project and project’s news. In addition through this area it is possible to submit to the coordinating center proposal for secondary analyses involving the use of the Shelter database
Achievements
Files with report on the Project Intranet Architecture (4 files), instructions for the use of the software, examples of CAP and PHP (A3) are available for examination by the commission.
Workpackage 4. Project protocol and Standard training definition
WP leader: partner 2 - CHARLES UNIVERSITY – FIRST MEDICAL FACULTY (CUNI)
Objective: to organise the Project pre-enrolment meeting and to plan and define a standardised training for the assessors who will collect data on residents (MDS LTCF) and on NH Services characteristics (EU-NHS).
Status of the WP: completed
During the pre-enrolment meeting the project protocol was defined and standardization of training procedure was discussed. In addition partners were introduced to the EU-NHS form (Task 4.1). The discussion of the project protocol was focused on the following key issues:
ENROLMENT AND FU TARGET POPULATION
Overall 500 participants will be enrolled in each site. Based on the project protocol residents should be re-assessed at 6 and 12 months. Given the necessity to reassess patients an effort is required to study partners to oversample long stay residents.
RELIABILITY DATA
Inter-rater reliability: InterRAI LTCF will be administered to 50 residents by two independent raters in each country (possibly in the same day).
Test-retest reliability: InterRAI LTCF will be administered to 50 residents on two different occasions (possibly the 2 assessment should be performed within 48 hours, 72 hours at the most).
It is key the health conditions of the resident are not changed between the 2 assessment.
DISCHARGE ASSESSMENT
Residents assessment will be performed at 6 and 12 months. An additional assessment will be filled-in when the participant is discharged from the facility, hospitalized or dead. In case of death it will not be necessary to fill in the whole assessment, but only a part of it, indicating the date of death
As already mentioned also standardization of training was discussed and standardized (Task 4.2). Nationally, courses were be organised to:
• Give nurses theoretical bases in the main gerontological and geriatric problems, along with instructions on the concept of comprehensive geriatric assessment and multidisciplinary teamwork;
• Train nurses to the use of the MDS LTCF;
In the first part of the course main gerontological and geriatric problems should be presented through classes, seminars and conferences. The concept of comprehensive geriatric assessment and multidisciplinary teamwork will be introduced and multidimensional assessment tools will be presented.
In the second part of the course participants should learn how to perform MDS assessment, including the specific forms and appropriate coding. They must also develop a comprehensive understanding of the connection between MDS and the Care Planning Process. Strategies for effective care planning must be discussed. Teachers will carefully explain:
1. The intent of the listed in the MDS LTCF, explaining criteria for inclusion, discussions of how the information will be used by clinical staff to identify person problems and develop care plan.
2. Supplemental definitions, examples, and instructions for completing MDS LTCF items.
3. Reminders of which MDS LTCF items require observation of the person for other than the standard 3-day observation period.
4. Sources of information to be consulted in completing specific MDS LTCF items.
5. Coding, explaining proper method of recording each response, with explanations of individual response categories. Coding of the following items will be analyse in depth: Identification Information, Demographic information, Cognitive patterns, Communication/hearing patterns, Vision patterns, Abuse and neglect, Mood and behaviour patterns, Psychosocial well-being, Physical functioning and structural problems, Continence, Disease diagnoses, Health conditions, Oral/nutritional status, Oral/dental status, Skin condition, Activity pursuit patterns, Medications, Special treatments and procedures, Discharge potential and overall status, Assessment information, Medications, Resident assessment protocol summary.
A simulated case must be presented and discussed in group.
Audiovisuals may be used including transparencies, slides and videotapes to present simulated Services and Health for Elderly in Long TERm care cases. At the end of the courses a final evaluation of participating nurses will be performed.
Participating nurses should also be invited to evaluate teachers and write their feedbacks on specifically designed anonymous forms.
Additionally a training course must be organized to train research assistants to the use of the EU-NHS form. During this course the EU-NHS must be explained and practical example will be provided.
MDS LTCF manual (English version) was made available for research nurses training and translated by participated partners if necessary (see file MDS LTCF manual).
Achievements
Files on Shelter Project final protocol, guidelines on training, pre-enrolment meeting
Minutes, MDS LTCF manual (A4) and one page summary of the project, posters for a non specialist audience (A5) are available for examination by the commission.
Workpackage 5. Training
WP leader: partner 1 – UNIVERSITA’ CATTOLICA DEL SACRO CUORE (UCSC)
Objective: to train research nurses to perform the assessment and data collection.
Status of the WP: completed
Courses to train research nurses to collect data on residents enrolled in the study were organised locally (Task 5.1). Training was performed by participating partners locally following guidelines discussed during the pre-enrolment meeting. The training was based on a 5-days, 5 hour a day course. The course consists of two parts, the first one on the basic concepts of gerontology and geriatrics and on the theories of geriatric nursing; the second is a practical training where the participants learn the use of the Minimum Data Set. Both sections are based on the MDS – LTCF. This course has been tested before, and the results obtained have been published on international journals (The MDS instrument as teaching tool for nursing home staff, Educat Geront 1996; Variation in training programs for Resident Assessment Instrument implementation, Age Ageing 1997; A brief training program on RAI instrument improves motivation of nursing home staff, Fact Research Gerontol 1998). The details of the classes are as follows:
1st day: Introduction to the course, principal theories on aging, demographic and epidemiological aspects of aging, multi-dimensional assessment and principal multi-dimensional assessment tools;
2nd day: Introduction to the Minimum Data Set – Long Term care Facility, analysis of the items of the sections AA through H, practical training on the evaluation of the Activities of the Daily Living;
3rd day: Analysis of the items of the sections H through Q, practical training on the whole assessment form, introduction to the Client Assessment Protocols;
4th day: Use of the Minimum Data Set to identify and evaluate the principal geriatric syndromes: delirium, cognitive impairment/dementia, communication/hearing, vision, physical functioning / activities of daily living, instrumental activities of daily living;
5th day: Use of the Minimum Data Set to identify and evaluate the principal geriatric syndromes: rehabilitation potential, bladder continence / indwelling catheter, psychosocial well-being, mood status.
At the end of the course, the participants were able to perform a multi-dimensional evaluation of the client using the Minimum Data Set. They were also be able to design an individual care plan allowing to maximize and/or maintain the best possible level of physical functioning and psychosocial well-being. Simulated cases were used to evaluate information retained from the course (Task 5.2).
Locally research assistant were trained to the use of the EUNHS form in a 1-day course (Task 5.3). For each participating facility a research assistant responsible for collecting data of the EUNHS form was identified.
Achievements
Internal report on procedures for nurse training (IR1) is available for examination by the commission.
Workpackage 6. Residents enrolment and follow up
WP leader: partner 8 – UNIVERSITY OF KENT AT CANTERBURY (UniKent)
Objective: to standardise the procedures and to start the enrolment phase.
Status of the WP: completed
Since the MDS-LTCF instrument is supposed to be used to assess nursing home residents irrespective of age, diseases and health condition, no specific exclusion criterium was used to select residents participating to the study. Only inclusion criterium was being a nursing home resident. In addition, partners were suggested to oversample, if possible, long stay residents because a 6 and 12 months follow up was planned for the project.
Residents were enrolled from different nursing homes in participating countries. Every nursing home in the project must be registered in the website and EUNHS form completed.
Finally, since the aim of the SHELTER project is to validate a methodology to analyze the provision of care in nursing home in Europe and not to compare characteristics of NH residents and systems it was decided not to randomize facilities and patients in order to have a country representative sample.
Resident enrolment has started at month 10 (October 2009). Overall, 4156 residents from participating countries (original target was 4000 residents) were enrolled and completed the baseline assessment (Task 6.1 and 6.2). In addition, 784 residents were enrolled to calculate test-retest reliability and inter-rater reliability. Test-retest reliability was evaluated in 380 residents by 48 assessors (nurses: 42; MD: 5, psychologist: 1) in 17 facilities (Task 6.3). This sample was composed of 52 dual assessments from Italy, 50 from Czech Republic, Germany and Israel, 48 from England and France, 43 from Finland and 39 from the Netherlands. Inter-rater reliability was evaluated in 404 residents by 76 assessors (nurses: 68; MD: 4; psychologists: 4) in 23 facilities (Task 6.4). This sample was composed of 60 dual assessments from Israel, 59 from Italy, 50 from England, 49 from France and Czech Republic, 48 from Germany and Finland and 41 from the Netherlands. The baseline assessment was completed at month 13 (December 2009). Six and twelve months follow up assessments were completed by December 2010 (Task 6.6 and 6.7). In addition, residents leaving the facility before the end of the study were reassessed by study staff (discharge assessments). Number of residents completed the baseline, follow up and discharge assessments by country are listed in the table 1 (see attached documents).
During the follow up data on mortality, including date, place and cause of death are collected by study nurses as well as data on hospital admissions and length of stay, emergency room visits, hospital outpatient referral, and day care (Task 6.8).
Participating residents were enrolled in 57 facilities. For each facility the EUNHS form was completed and data collected through the study website (Task 6.5). Geographic distribution of participating facilities is presented in figure 1 (see attached documents).
Achievements
Procedures and criteria for patient enrolment (1 file - IR2) are available for examination by the commission.
Workpackage 7. Study monitoring
WP leader: partner 1 – UNIVERSITA’ CATTOLICA DEL SACRO CUORE (UCSC)
Objective: to ensure both standardisation and timely progress through all the study stages.
Status of the WP: completed
The appointed partner’s 1 (UCSC) staff member, with considerable experience in this field having already co-ordinated at a European level studies on implementation of comprehensive assessment instruments, visited the participating centres in order to review the progress in data collection, as well as to detect, discuss and solve with local partners any problem arising regarding residents and Services assessment, data interpretation, collection or storage (Task 7.1 and 7.2). Quality assurance was also ensured during site visits with the production of site Gantt chart to schedule all the foreseen activities. Partner’s 1 staff visited each study site twice every year during the phase of enrolment and follow up. All information from such visits were analysed by the Steering Committee in accordance with the procedures detailed in the workplan and in the management section (Task 7.3). Monitoring of the activities of participating facilities was performed locally by project partners. Reports on data collection progress and problem assessment were discussed during the mid term meeting and Steering Committee meetings. Tables reporting status of the enrolment procedures in different countries were provided to partners and constantly updated. The following problems were identified during the mid-term meeting:
Partner 4 (THL) did not provide the translated version of the InterRAI LTCF in doc format. No data on assessments were available in the software for the following partners: Partners 4 (THL), 5 (AP-HP.SPR) and 9 (VUA).
The partners were committed to complete their duties as follows:
Partner 9 (VUA): baseline and test-retes and inter-rater reliability data made available by end of March, 2010
Partner 5 (AP-HP.SPR): baseline and test-retes and inter-rater reliability data made available by end of May, 2010
Partner 4 (THL): translation of InterRAI LTCF and baseline and test-retes and inter-rater reliability data made available by end of April, 2010. Now these problems are fully solved and the project does not present any delay.
Achievements
Gantt charts to monitor study progress (IR3), Mid-term meeting report indicating problems on data collection progress and assessment (A6) and Steering Committee reports indicating problems on data collection progress and assessment (A7 and A8) are available for examination by the commission.
Workpackage 8. Shelter database implementation
WP leader: partner 9 – VERENIGING VOOR CHRISTELIJK HOGER ONDERWIJS WETENSCHAPPELIJK ONDERZOEK EN PATIENTENZORG (VUA)
Objective: to create the Shelter database, trough the collection of the following information: a. residents baseline and follow-up (6 and 12 months) data collected by the MDS LTCF; b. data on reliability; c. structural and organizational characteristics of the nursing home Services, collected by the EU-NHS form; d. data on mortality and health care resource consumption.
Status of the WP: completed
Residents-level data were collected at baseline, six and twelve months by research nurses trained on the use of MDS LTCF in each country. In addition, to evaluate reliability of the instrument dual, independent assessments will be performed in a sample of 100 residents in each country. Study researchers responsible for data entry were provided with password to access database and data entry (Task 8.1).
A variety of validation checks were applied during the data entry process (Task 8.2). These included intra-form and inter-form checks for consistency of responses, range checking, and checks for required fields and skip patterns. More sophisticated checks were made after the data have been entered. Errors and inconsistencies detected at this stage may generate queries for the clinic staff.
Data entry took taking place via an internet-based web browser interface (Task 8.3). Electronic data were managed centrally at the Co-ordinating Centre, with a feedback mechanism through e-mail and web-based reporting. Raw files were generated from this process (Task 8.4).
Achievements
First set of analytical files collecting data on baseline assessment, drug use and facility characteristics (A9), a second (A10), third (A11) and final set (A12) of analytical files are available for examination by the commission.
Workpackage 9. Data analysis
WP leader: partner 8 - UNIVERSITY OF KENT AND CANTERBURY (UniKent)
Objective: To analyse the reliability of items included in the MDS LTCF; the baseline demographic, clinical and functional characteristics of the Nursing Home residents in the selected countries; quality of care provided in the NH included in the study; resident’s outcomes, to identify predictors (both at a patient level and at a service level) of mortality, health resources consumption and other functional and clinical outcomes.
Status of the WP: completed
Analytical files were created for data analysis (Task 9.1) and the following analyses were performed:
Reliability analysis (Task 9.2)
Test-retest reliability and Inter-rater reliability of items included in the MDS LTCF were measured estimating Kappa statistics for nominal data (Kappa values below 0.4 were considered indicators of poor reliability, values between 0.4 and 0.75 were considered indicators of adequate reliability and values of 0.75 or above were considered indicators of excellent reliability). Pearson’s correlation coefficient (r) were calculated to measure agreement for continuous and ordinal data respectively.
Overall, reliability of 201 InterRAI LTCF items was tested. Only 3 items (walking speed, height and weight) were continuous variables. Table 2 (see attached documents) presents average weighted kappa values for categorical items in different assessment areas. Test-retest reliability was evaluated in 380 residents by 48 assessors (nurses: 42; MD: 5, psychologist: 1) in 17 facilities. This sample was composed of 52 dual assessments from Italy, 50 from Czech Republic, Germany and Israel, 48 from England and France, 43 from Finland and 39 from the Netherlands. As shown in table below, test retest reliability of InterRAI LTCF items in each of the assessment areas ranged from 0.75 to 0.92. Average weighted kappa for the 198 categorical items was 0.83. None of the single items in the InterRAI LTCF presented with a kappa value below 0.40. Pearson’s correlation coefficients for items based on continuous variables were 0.96 for walking speed, 0.92 for height and 0.96 for weight.
Inter-rater reliability was evaluated in 404 residents by 76 assessors (nurses: 68; MD: 4; psychologists: 4) in 23 facilities. This sample was composed of 60 dual assessments from Israel, 59 from Italy, 50 from England, 49 from France and Czech Republic, 48 from Germany and Finland and 41 from the Netherlands. As shown in table below inter-rater reliability of items in the assessment areas ranged from 0.64 to 0.91. Average weighted kappa for the 198 categorical items was 0.74. Only one single item had a kappa value below 0.40 (‘Fluid output exceeds input’ included in the Oral/Nutritional Status assessment area). Pearson’s correlation coefficients for items based on continuous variables were 0.88 for walking speed, 0.99 for height and 0.89 for weight.
Description of baseline characteristics of study sample (Task 9.3)
As mentioned, total number of 4156 residents was enrolled in the study, 500 from Czech Republic, 507 from England, 484 from Finland, 493 from France, 496 from Germany, 580 from Israel, 548 from Italy and 548 from the Netherlands. Table 3 (see attached documents) summarizes the principal characteristics of NH residents enrolled in the study. Mean age was above 80 years and women represented approximately 3/4 of the sample. Disability was common, with more than 80% of participating residents requiring assistance or being dependent in ADL and cognitive impairment was present in more than 2/3 of the sample with 30% of residents being classified as severely impaired. Elevated level of “clinical” complexity of residents was confirmed by the high prevalence of urinary incontinence (73.5%), pain (36.0%), depression (32.0%), behavioral symptoms (27.5%), falls (18.6%) and pressure ulcers (10.4%).
Analysis of quality of care (Task 9.4)
The MDS LTCF produces Quality Indicators (QIs) relevant for long term care. Some of these QI are summarized in table 4 (see attached documents).
In particular, some QI were analysed in depth. For example, an ‘adhoc’ analysis was dedicated to polypharmacy (concurrent use of 5-9 drugs) and excessive polypharmacy (concurrent use of ≥ 10 drugs) .
Polypharmacy was observed in 2000 (49.7%) residents and excessive polypharmacy in 979 (24.3%) residents. As shown in Figure 4 (see attached documents), prevalence of polypharmacy and excessive polypharmacy widely varied in study sample. In particular, Italy had the lowest prevalence of excessive polypharmacy (8.8%), followed by Israel (12.9%), Germany (15.7%), England (22.7%), the Netherlands (24.4%), Czech Republic (25.2%), France (30.2%) and Finland (56.7%).
As shown in table 5 (see attached documents) residents on polypharmacy and excessive polypharmacy, when compared with those not on polypharmacy, had a less severe level of disability and cognitive impairment, a reduced rate of behavioural symptoms and a higher prevalence of depression, falls, pain, dyspnoea, GI symptoms, dizziness and flare up of a recurrent or chronic problem. Finally, residents on polypharmacy and excessive polypharmacy presented a higher number of concomitant diseases and, in particular, ischemic heart disease, heart failure, Parkinson disease, stroke, diabetes and cancer were more common in these groups as compared with the non-polypharmacy group.
Table below reports results of the multivariate analysis, identifying variables independently associated with polypharmacy and excessive polypharmacy. As compared with non-polypharmacy, excessive polypharmacy was directly associated with depression (OR 1.81; 95% CI 1.38-2.37) pain, dyspnoea, GI symptoms, and specific diseases including ischemic heart disease, heart failure, Parkinson disease, stroke and diabetes, while an inverse association was shown for age, ADL disability and cognitive impairment. Similarly, polypharmacy was directly associated with depression, pain, ischemic heart disease, heart failure, Parkinson disease and diabetes. Presence of behavioural symptoms presented an inverse and significant association with polypharmacy, but not with excessive polypharmacy.
Facilities characteristics - Resources utilization (Task 9.5)
Characteristics of participating facilities were analysed using data from the EUNHS form. These data are summarized in table 6.
MDS LTCF assessment systems support the use of algorithms that group individuals into categories reflecting the relative costs of services they are likely to consume. These tools are designed to address an enormous challenge faced by governments the world over: how to allocate limited resources to persons in need in a fair and equitable way. Case mix is by definition a system that classifies people into groups that are homogeneous in their use of resources. A good case-mix system also gives meaningful clinical descriptions of these individuals. The application of case mix is broad; it provides the basis, not only for reimbursement, but also for comparing facilities or programs, practice patterns, as an adjunct to quality of care and efficiency measurement, a staff planning tool, etc.
The Resource Utilization Groups (RUG-III) case-mix algorithm was developed to provide a patient-specific means of allocating health care resources based on the variable costs of caring for individuals with different needs. The RUG-III uses 108 variables from the MDS LTCF to create 44 34 categories of patients with homogeneous resource use patterns. The RUG-III algorithm explains about 55% of variance in resource use, and it has been validated in a number of countries through a series of international studies. Figures 5 and 6 (see attached documents) present the RUG-III distribution of residents in the SHELTER study.
Identify independent predictors of residents outcomes (Task 9.6)
Participating residents were followed for 1 year. Of the 4156 residents in the baseline samples, follow up data were collected for 3741 residents (drop out rate 10%). Overall 761 residents died during the follow up (20.3%). Cox analysis was perform to identify independent predictors of mortality. Results of this analysis is presented in the table 7 (see attached documents).
Workpackage 10. Dissemination of project results
WP leader: partner 3 - UNIVERSITY OF ULM (UUlm)
Objective: to disseminate the Project’s results
Status of the WP: completed
The dissemination of projects results was tackled in different ways:
1. Organization in symposia during scientific meetings (Task 10.1):
a. A symposium was organized during the VII EUROPEAN INTERNATIONAL CONGRESS of the INTERNATIONAL ASSOCIATION OF GERONTOLOGY AND GERIATRICS (IAGG) - HEALTHY AND ACTIVE AGEING FOR ALL EUROPEANS held in Bologna (Italy), April 14-17, 2011. The Symposium entitled ‘LONG-TERM CARE CHALLENGES IN EUROPE - STANDARDIZED ASSESSMENT FOR ADDRESSING PATIENT’S NEEDS, OUTCOMES AND QUALITY OF CARE’ included four talks focused on results of the SHELTER study;
b. During the 2008 meeting of the Italian Society of Gerontology and Geriatrics (SIGG) a special symposium on ‘EUROPEAN RESEARCH PROJECTS ON HEALTH SYSTEMS AND LONG TERM CARE OF THE ELDERLY’ was held. The symposium involved representatives from the Shelter project as well as investigators from other projects funded under the FP7, the Interlinks project and the Ancien project;
c. During the 2010 ECHE meeting in Helsinki (FI), during a session on ‘Long-Term Care for the Elderly: A European Perspective’ the design, methodology a status of the Shelter project were presented.
2. Submission of manuscripts to scientific peer reviewed journals (Task 10.2). So far the following manuscripts based on SHELTER data were published:
a. Onder G, Carpenter I, Finne-Soveri H, Gindin J, Frijters D, Henrard JC, Nikolaus T, Topinkova E, Tosato M, Liperoti R, Landi F, Bernabei R. Assessment of nursing home residents in Europe: the Services and Health for Elderly in Long TERm care (SHELTER) study. BMC Health Serv Res. 2012 Jan 9;12(1):5.
b. Onder G, Liperoti R, Fialova D, Topinkova E, Tosato M, Danese P, Gallo PF, Carpenter I, Finne-Soveri H, Gindin J, Bernabei R, Landi F; for the SHELTER Project. Polypharmacy in Nursing Home in Europe: Results From the SHELTER Study. J Gerontol A Biol Sci Med Sci. 2012 Jan 4.
c. Tosato M, Lukas A, van der Roest HG, Danese P, Antocicco M, Finne-Soveri H, Nikolaus T, Landi F, Bernabei R, Onder G. Association of pain with behavioral and psychiatric symptoms among nursing home residents with cognitive impairment: Results from the SHELTER study. Pain. 2012 Feb;153(2):305-10
3. The final conference of the project entitled ‘Long Term Care of Europe’s Older Citizens: Policy and Fact. Providing data to support the relationship between care practice organization and quality of care in long term care facilities in Europe’ that was held in Brussels on December 5, 2011 constituted “per se” a form of dissemination activity. It was attended by the project partners and by stakeholders, national policy makers, health administrators and politicians responsible for health decisions. This meeting constituted a way to make many opinion leaders in the health services for the elderly field aware on relevant areas of policy related to SHELTER findings and illustrate the policy problems for providing high quality care for frail older people in NH.
4. A one page promotional summary of the project was produced to announce its aims and generate interest among a broad audience. A more detailed overview was provided in a project identity brochure, drawing on background information from the SHELTER project.
5. A website (www.shelter-elderly.eu) was established and regularly updated with the newest information about the project. The SHELTER website describes the aims of the project and collected all relevant documents related to the project, including the final conference program, the promotional summary the brochure and the slides presented in the final meeting.
Achievements
Program of the symposium held at the meeting of the INTERNATIONAL ASSOCIATION OF GERONTOLOGY AND GERIATRICS (IAGG – A18), as well as programs of the meeting of the Italian Society of Gerontology and Geriatrics and of the ECHE meeting, papers accepted for publication (A19), the program of the final conference, the brochure and the flyer are available for examination by the commission.
Potential Impact:
Impact
The present project validates a new methodology for analysing provision of long term care in Europe. To date, no comprehensive assessment tool specifically targeting European long term care industry has been developed and validated. As such, this is the first attempt to demonstrate the efficacy of a scientific methodology for understanding current services and care processes in European nursing homes, addressing the needs of individual residents and providing evidence for resource allocation, planning and organization.
The ageing of the population is leading health care expenditure to be one of the greatest socioeconomic issues for European countries. The ageing phenomenon will be responsible for a quite dramatic societal change with less working age individuals available to take care of an instead progressively larger number of disabled and dependent elderly individuals. Under this scenario, the long term care industry is projected to rapidly expand with a substantial impact on public spending and on planning of health care services by the governments of the States Members.
This foreseeable public health emergency requires a strategic plan using a scientific approach.
The main barrier for an efficient public health intervention in this field is represented by the nearly total lack of knowledge about long-term care services across Europe. Information about structural and organizational characteristics of long term care services is rather sparse, and documentation extremely heterogeneous and influenced by cultural and societal expectations. In those countries where the system is regulated, extremely different and not comparable standards are adopted.
This applies to eligibility criteria, assessment instruments, protocols and procedures, financing mechanisms. Thus, the implementation of a standardized, comprehensive, cross-languages and cross–cultures assessment instrument is believed to be a crucial, initial step toward the acquisition of the necessary knowledge base.
The current project validates the MDS tool as the first comprehensive assessment instrument, specifically intended for the application within the European nursing home system. Scientific evidence was produced to nursing home operators, medical directors, administrators, policy makers, public health experts, health care professionals and researchers that the MDS represents a valid and readily applicable instrument. A large scale implementation across Europe could now be promoted. The systematic collection of information with the MDS instrument led to the creation of a cross-national database including data on nursing home residents and services characteristics. A systematic collection of data using the MDS tool provided reliable and detailed cross-national data on nursing home industry in Europe. For the first time, countries share a common language to understand, analyse and manage what is going to be the most critical sector of the health care industry in the near future.
Data sharing gets European countries to learn from each other and to manage diversity and heterogeneity in respect of a unique high standard of quality and cost-effectiveness. Such methodology produces convincing evidence concerning clinical outcomes of individuals or groups of residents, eligibility criteria, to ensure monitoring of services delivery, to measure quality and to estimate efficacy and effectiveness of prevention programs. Also, costeffectiveness of different health care strategies from different countries were scientifically compared and the best model of care may be identified. An ultimate impact on the reduction of public spending for long term care and on the rationalization of resources allocation in health care is much expected.
Dissemination
The dissemination of projects results was tackled in different ways:
1. Organization in symposia during scientific meetings:
a. A symposium was organized during the VII EUROPEAN INTERNATIONAL CONGRESS of the INTERNATIONAL ASSOCIATION OF GERONTOLOGY AND GERIATRICS (IAGG) - HEALTHY AND ACTIVE AGEING FOR ALL EUROPEANS held in Bologna (Italy), April 14-17, 2011. The Symposium entitled ‘LONG-TERM CARE CHALLENGES IN EUROPE - STANDARDIZED ASSESSMENT FOR ADDRESSING PATIENT’S NEEDS, OUTCOMES AND QUALITY OF CARE’ included four talks focused on results of the SHELTER study;
b. During the 2008 meeting of the Italian Society of Gerontology and Geriatrics (SIGG) a special symposium on ‘EUROPEAN RESEARCH PROJECTS ON HEALTH SYSTEMS AND LONG TERM CARE OF THE ELDERLY’ was held. The symposium involved representatives from the Shelter project as well as investigators from other projects funded under the FP7, the Interlinks project and the Ancien project;
c. During the 2010 ECHE meeting in Helsinki (FI), during a session on ‘Long-Term Care for the Elderly: A European Perspective’ the design, methodology a status of the Shelter project were presented.
2. Submission of manuscripts to scientific peer reviewed journals. So far the following manuscripts based on SHELTER data were published:
a. Onder G, Carpenter I, Finne-Soveri H, Gindin J, Frijters D, Henrard JC, Nikolaus T, Topinkova E, Tosato M, Liperoti R, Landi F, Bernabei R. Assessment of nursing home residents in Europe: the Services and Health for Elderly in Long TERm care (SHELTER) study. BMC Health Serv Res. 2012 Jan 9;12(1):5.
b. Onder G, Liperoti R, Fialova D, Topinkova E, Tosato M, Danese P, Gallo PF, Carpenter I, Finne-Soveri H, Gindin J, Bernabei R, Landi F; for the SHELTER Project. Polypharmacy in Nursing Home in Europe: Results From the SHELTER Study. J Gerontol A Biol Sci Med Sci. 2012 Jan 4.
c. Tosato M, Lukas A, van der Roest HG, Danese P, Antocicco M, Finne-Soveri H, Nikolaus T, Landi F, Bernabei R, Onder G. Association of pain with behavioral and psychiatric symptoms among nursing home residents with cognitive impairment: Results from the SHELTER study. Pain. 2012 Feb;153(2):305-10
3. The final conference of the project entitled ‘Long Term Care of Europe’s Older Citizens: Policy and Fact. Providing data to support the relationship between care practice organization and quality of care in long term care facilities in Europe’ that was held in Brussels on December 5, 2011 constituted “per se” a form of dissemination activity. It was attended by the project partners and by stakeholders, national policy makers, health administrators and politicians responsible for health decisions. This meeting constituted a way to make many opinion leaders in the health services for the elderly field aware on relevant areas of policy related to SHELTER findings and illustrate the policy problems for providing high quality care for frail older people in NH.
4. A one page promotional summary of the project was produced to announce its aims and generate interest among a broad audience. A more detailed overview was provided in a project identity brochure, drawing on background information from the SHELTER project.
5. A website (www.shelter-elderly.eu) was established and regularly updated with the newest information about the project. The SHELTER website describes the aims of the project and collected all relevant documents related to the project, including the final conference program, the promotional summary the brochure and the slides presented in the final meeting.