Growing old at home – A randomized controlled trial to investigate the effectiveness and cost-effectiveness of preventive home visits to reduce nursing home admissions: study protocol [NCT00644826]

The study is conducted in cooperation with the university hospital Halle, the Diakoniekrankenhaus Halle, the registration office Halle and 3 GP practices in the city of Halle and 18 GP practices in the city of Leipzig. All people older than 80, fluent German speakers, residents of Leipzig or Halle, living at home or planned discharge to home (hospital patients) are eligible for this study. Additionally, participants have to be impaired in at least 3 activities of daily living. We exclude from trial participation people that are cognitive impaired, not able to give informed consent or have a care level higher than I (according to German long term care insurance).

It is a well documented problem to recruit participants in this age group especially when interventions that demand home visitation are involved [11]. We therefore will implement three strategies to maximize recruitment and reach the needed sample size. The primary recruitment strategy is via GP practices in the cities of Halle and Leipzig. As a second strategy the recruitment via liaison nurses will be implemented in hospitals in Halle to cope with the small number of participating GPs and to include a group of high risk participants. As a third recruitment strategy people aged 80 and over in Halle and Leipzig will be contacted by mail. Addresses will be retrieved by cooperating with the local registration offices. We will send explicit trial information, a consent form for participants and a self-addressed stamped envelope. After return of the consent form we will contact the responders by telephone, prescreen for inclusion and exclusion criteria and make an appointment for a home visit. This first home visit is different to the first home visits of the other recruitment methods so far that the screening is done within that visit in addition to the assessment. Prescreening and final screening is necessary as we expect a certain amount of inappropriate responders to our letter. We will set a time limit of 4 weeks after dispatch of the last letter till we close participant registration via mail. All received consent letters then will be put in random order. This will be the order we contact the persons until we got enough study participants in addition to general practitioners' practices and hospital recruitment. Exceeding persons will get a written cancellation due to numerical limitation for trial participants.

A balanced block-wise randomization, stratified by center, will be used. In Halle the sample additionally is stratified according to the recruiting hospital and practice. Concealment of allocation is assured by central randomization via consecutive randomization lists in the order of recruitment.

Participants recruited via the registration office will be randomized using sealed opaque envelopes immediately after completion of the baseline assessment.

Patients and field researchers cannot be blinded to group allocation. A blinded statistician not involved in trial conduct will do the final analysis.

A liaison nurses screens and recruits the participants for eligibility in the 2 hospitals in Halle. Medical secretaries screen and recruit the participants in Leipzig and Halle for eligibility in the participating GPs' practices. Patients then are randomized by the local study center via consecutive randomization lists in this way assuring concealed allocation of participants. The participants are contacted by the centers' study personal via mail and phone. In the phone call an appointment for the first home visit is made with the patients. In Halle we additionally will contact the patients after discharge from hospital. All other study procedures are standardized across the study centers, in this way ensuring a homogenous conduct of the trial.

The intervention group will be assessed with a geriatric assessment and the baseline assessment. A team of nurses, psychologists, physiotherapists, dietitians and physicians will work out individualized recommendations in a case review one week after the geriatric assessment. The main aim is to identify self-care deficits and risks for self-care deficits in the socioeconomic supportive dimension, the social integrative dimension and finally the health dimension. Then appropriate recommendations will be given by the team. After the case conference the members of the intervention group are visited by the same staff member again. This staff member will conduct a consultation and instruction based on the results of the case conference. Four weeks after the consultation another visit is done by the staff member. This visit (booster session) is for evaluative purposes mainly: To what extent the study participant adheres to the recommendations and how convenient are the recommendations. The final measurement will be the follow-up assessment 18 months after baseline.

The control group is only assessed with the baseline assessment and the follow-up assessment 18 months later. The control group will not receive any intervention in addition to standard care. The flow of the participants is shown in Figure 1.

Primary outcome of our trial is the incidence of nursing home admission over the study period of 18 months. Nursing home admission in our trial is defined as the permanent admission into an inpatient nursing care facility according to the social security code of Germany. Short term care, assisted living, geriatric day care or respite care facilities are not included in this definition.

Data will be collected on follow-up assessment either by direct contact with the participants, their relatives or the resident registration office.

Secondary outcomes over the whole trial period of 18 months:

Sample size calculation was based on an assumed absolute risk reduction from a 20% in the control group to a 7% in the home admission rate of the intervention group. Power calculation (α = 0.05; β = 0.20) including an estimated drop out rate of 30% resulted in a required sample size of N = 320 (n = 160 vs. n = 160) using a two-tailed Chi-square test. The sample size is equally divided among the two study centers Halle and Leipzig.

Drop-outs will be documented thoroughly and included in data analysis to the point of drop-out. Reasons for drop-out will be reported and analyzed.

Clinical data analysis will be done by the institute for epidemiology, biostatistics and medical informatics of the medical faculty Martin-Luther-University Halle-Wittenberg.

In a first step adequate descriptive statistics will be used to compare the intervention group against control group after randomization. Then all outcomes will be tested in comparison of the intervention and control group with multifactorial regression models on an intention-to-treat basis. Level of significance is determined by 5%. Two-tailed tests for significance will be used for all statistical tests.

Precision of results will be specified. Confidence intervals (95%) will be computed for all primary and secondary outcomes.

Economic analysis will be conducted by the Health Economics Research Unit, University of Leipzig. This involves calculating costs as well as the cost-effectiveness ratio, i.e. the ratio of the difference in mean costs and the difference in mean effects between the intervention and control group. QALYs based on EQ-5D [12] will be used to measure health effects in cost-utility analysis. To assess the uncertainty of the results sensitivity analysis will be performed and cost-effectiveness acceptability curves will be computed.

The trial is part of the Nursing Research Network "Mitte-Süd". A report system is established within the network. Annual quality reports have to be prepared for the German Federal Ministry of Education and Research.

As the conduct of a multi center trial demands high standards of quality to warrant comparable conditions and results among the centers all procedures were developed and documented in joint commissions.

Screening includes sociodemographic data (age, sex, marital status and housing conditions), information about long term care insurance, nursing allowance and degree of impairment to receive benefits from the long term care insurance. Furthermore, Activities of Daily Living (ADL) are checked [13].

All participants (intervention and control group) receive baseline assessment after randomization and the same assessment at follow-up after 18 months. Cognitive function is assessed using the Mini Mental State Examination (MMSE) [14]. The health related quality of life is measured by the EuroQol-5D [12], including a visual analogue scale (VAS) to measure patients preferences. Psychosocial factors are assessed with the Social Situation by Nikolaus (SoS) [15]and the 5-item version of the Geriatric Depression Scale (GDS)[16]. This short form was chosen because it is as effective as the 15-item version of the GDS [17], which is widely used for depression screening in cognitively unimpaired elderly persons. Moreover, the 5-item GDS is more an economical version because of reduced administration time. Details about the functional status will be determined using the Barthel-Index [18] and the Instrumental Activities of Daily Living (IADL) [19]. Additionally, all participants are asked about their history of falls in the past 12 months. A Questionnaire of Service Utilization and Costs will be used based on cost diaries used in earlier studies [20‐25].

This questionnaire also includes sociodemographic data (i.e. age, sex, marital and educational status, life conditions); information about health insurance, long term care insurance and the degree of impairment to receive benefits from the long term care insurance. Furthermore the Chronic Disease Score will be calculated [26].

All subjects of the intervention group additionally receive intervention assessment immediately after baseline assessment. The current nutrition status is measured by the Mini Nutrition Assessment (MNA) [27]. Other health dimensions of elderly persons such as impaired sight or hearing, urinary or bowel incontinence and loss of functional muscle mass are assessed with the Geriatric Screening AGAST [28] and the Geriatric Screening by Lachs [29]. These geriatric screenings also determine social activities, housing conditions, economical conditions and polypharmacy. The Clock-Completion Test [30] will complement the MMSE for the baseline assessment of cognitive abilities.

We will conduct interdisciplinary case conferences for all patients in the intervention group. We intend to collect several cases for a conference. We will conduct case conferences in the defined period of 3 weeks between first and second home visit in this way trying to standardize the procedures. Cases will be prepared and introduced to the conference group by the visiting investigator. The investigators will invite appropriate health professionals as the case requires. The professional spectrum of the conference expert advisory group hereby consists of nurses, a general practitioner, a nutritionist, a geronto psychiatrist, a physiotherapist, a psychologist and a social worker. Recommendations will be worked out for the single patient and we will document them for the next home visit. The duration of the preparation, present experts and duration of the conference will be documented for each patient individually.

The last part of the experimental intervention consists of a counseling intervention. We will address the identified problems and present the recommendations of our expert advisory group during this home visit. We estimate the duration of the second home visit including the counseling with 20–40 minutes.

Four weeks after the counseling home visit we will visit the patients in the intervention group for a third time. During this visit we want to evaluate which recommendations the patients already implemented and which not. In this way we can assess obstacles and facilitators to the recommendation adherence. Additionally, we have the opportunity to boost our recommendations from the second home visit.

18 months after the first home visit we will assess all participants using the same measurement as for the baseline assessment. Additionally, details about current residence status, nursing home admission and date of nursing home admission are included. If a subject cannot be reached by usual means like telephone or mail, the named contact person or residents' registration office will be contacted.

All measurements are summarized in Table 1.

We will create a pseudonym for all trial participants to collect and analyze the trial data. Key lists will be stored separately from the trial data and erased after final data analysis. Data will be analyzed in a way that no conclusions can be drawn to individual participants. Trial data is stored in lockable cabinets in lockable rooms. All gained address data of persons not included in the trial will be erased after exclusion (prescreening or screening) or non-inclusion (non-responders of the registration offices' samples).

The study protocol is approved by the ethics committees of the universities in Halle and Leipzig. If changes to the study procedures are necessary they will be proposed to the local ethics committees as amendments. All changes will be described and discussed in the publication of the trial's results.

We plan to publish the trial results in a peer reviewed international Medline-listed journal whether the effectiveness was shown or not. This mainly serves the purpose to avoid publication bias. Additionally, we are obliged by the Federal Ministry of Education and Research to report our results within 6 months after study termination. All trial results will be reported within context to this study protocol.