PReventing and Approaching Crises for frail community-dwelling patients Through Innovative Care (PRACTIC): protocol for an effectiveness cluster randomised controlled trial

Worldwide, the proportion and absolute number of older individuals are increasing dramatically. The population aged 60 and older is expected to double by 2050 worldwide, and the proportion of people receiving care at home has increased over the past 10 years [1, 2]. In Norway, approximately 200,000 people are currently receiving home care services, while there are approximately 40,000 beds in nursing homes [3]. The majority of people would rather reside and receive care at their homes than in an institution [4]. One of the main health and societal challenges for municipalities, now and in the future, is to offer high-quality health care services for a growing population of older people with complex needs due to frailty and multimorbidity [5]. Patients may experience distress because they are unable to manage the situation at home, even though they may prefer to avoid institutionalisation. The considerable distress experienced by informal caregivers revolves around the lack of sufficient support from home care services and the limited availability of nursing home places [6]. Home care services can be described as a complex organisation with various service components ranging from practical assistance in the home to the delivery of advanced medical treatment. To enable people to live safely at home, home care services are interdependent on other sectors, such as general practitioners (GPs), the hospital sector, primary health care workers, and social care [4, 7].

There is a significant variation among patients receiving home care services in terms of functional abilities, age, living conditions, and chronic diseases [8]. A considerable number of these patients have multiple chronic conditions, commonly referred to as multimorbidity [4]. Estimates suggest that within the next 20 years, the population of elderly individuals with multimorbidity will double [9]. The prevalence of frailty is estimated to be 11% among adults aged ≥ 65 years, increasing to 50% among those > 80 years of age [10]. Most frail individuals are multimorbid, but not all multimorbid individuals are frail [11]. Frailty has been described as a state of physiological vulnerability with a reduced capacity to adapt and manage internal and external stressors [12]. Studies have emphasised the relevance and utility of a biopsychosocial definition of frailty, including the terms physical frailty, psychological frailty, and social frailty. Social frailty is an important dimension of the frailty concept and makes people with low incomes, low educational levels, and low housing standards vulnerable to various adverse health outcomes. Combining these three dimensions into a multidimensional concept of frailty promotes the use of targeted multidomain interventions tailored to older adults’ frailty status [13, 14].

People who are at an increased risk of developing crises are often frail. Crises are major stressors for patients, their next of kin, and the care staff and often lead to adverse events, acute institutionalisation, and the use of coercion [15, 16]. Crises can be described as ‘a process in which the stressors cause an imbalance requiring an immediate decision which leads to a desired outcome and therefore crisis resolution’ [15]. In the Preventing and Approaching Crises for Frail Community-dwelling Patients Through Innovative Care (PRACTIC) study, we will operationalise this definition to describe crises in practice as ‘critical challenges and symptoms that demand immediate and new actions’. The challenges and symptoms that trigger and maintain crises are heterogeneous and vary between patients and may include depression, poor nutrition status, rejection of care, incontinence, neuropsychiatric symptoms (NPSs), and social isolation [12]. The ‘Mind the Gap report’ from the Advisory Board of the Global Forum for Health Care Innovators states that 1–5% of community-dwelling patients are high-risk patients and 15–35% are patients with an increasing risk. The literature on crises among patients receiving home care services has mainly explored the phenomenon in relation to people with dementia living at home [15, 16].

One of the most demanding challenges for health care authorities and home care services is to develop and implement high-quality health care models for the growing population of frail community-dwelling patients [17]. In addition to early recognition and response to clinical signs and symptoms, as recommended by the Norwegian Health Directorate, providing health care for this group of frail patients represents a change from a merely task-oriented service to a service that aims to assess the complex biopsychosocial character of frailty [18]. There are large variations in the content and organisation of Norwegian home care services, and research pertaining to these services has largely been descriptive, with a preponderance of qualitative studies [19, 20]. There is a paucity of studies investigating the effectiveness of interventions [20]. A study testing a structured follow-up programme using a checklist for frail community-dwelling adults found no common perception among nurses or their leaders that the approach was useful to ensure high-quality health care [21]. This finding supports the conclusion of a Cochrane Review in 2016 summarising primary care interventions for patients with multimorbidity [22]. The review revealed no clear positive improvements in clinical outcomes, health service use, medication adherence, patient-related health behaviours, health professional behaviours, or costs. The authors concluded that to improve outcomes for people with multiple conditions, there is a need for new multicomponent interventions, targeting both the heterogeneity of patients and their multimorbidity. To our knowledge, no effectiveness study of interventions targeting heterogeneous groups of home-dwelling patients with multimorbidity has been conducted in Norway. The proposed project will develop knowledge beyond the current state by also including the experimental testing of an intervention [20].

Participatory action research (PAR) in combination with an RCT has been suggested as a design to enhance local adaptations of an intervention to the local context and needs [23, 24]. There are multiple variations in the content and organisation of Norwegian home care services, with various service components ranging from practical assistance in the home to the delivery of advanced medical treatment [19]. The possibility of success for innovative interventions is probably higher if the interventions are not too complex, with no aim of changing the organisation of health care services [23, 25]. A flexible complex intervention has been emphasised as an important factor for interventions to be effective and increasing their applicability [24, 26‐28]. According to Hawe et al. [24], it is the function and processes of the intervention that should be standardised, not the components in the intervention. Adaptation of the components should be performed both at the research project level and at the implementation level in municipalities [24]. Using a PAR design will help to adapt the components of the intervention to these variations, thereby enhancing implementation in each municipality. As a part of the main PRACTIC study, a process evaluation study will be conducted in parallel with the RCT [29]. This will ensure that these variations in the organisation mentioned earlier and the necessary adaptations are accounted for.

TIME (Targeted Interdisciplinary Model for Evaluation and Treatment of Neuropsychiatric Symptoms) is a Norwegian evidence-based model for problem-solving regarding neuropsychiatric symptoms (NPSs) in dementia and other mental diseases. The model is based on the theoretical frameworks of cognitive behavioural therapy (CBT) and person-centred care (PCC) [28]. TIME has also been used in clinical practice for other complex issues, such as nutritional failure, multimorbidity, and general functional loss [30]. It is a multicomponent interdisciplinary model consisting of three overlapping phases, which are the core components of the model. First is the assessment phase where the care staff and the physician collaborate in a comprehensive biopsychosocial assessment. The second phase is the reflection phase with interdisciplinary case conferences based on principles from cognitive behavioural therapy (the ABC method), where a customised treatment plan is developed. The ABC method from cognitive behavioural therapy is used as an analytic tool for the analyses of complex challenges in case conferences [31]. The third phase is the action and evaluation phase, and each treatment measure in the plan is implemented and systematically evaluated. The TIME model is effective for treating NPSs in dementia and has been proven feasible in nursing homes (NHs) [28, 32]. Our research centre has pilot tested the model in home care services [33]. One of the assets of TIME is interdisciplinary case conferences, and interdisciplinarity is essential in the approach to a crisis. Based on the results from the pilot test, the inclusion criteria for the patients were broadened, and the content of the training for all employees was further developed [33]. In addition, the schedule for the training and implementation was adapted to everyday routines for home care services. Further adaptation of the TIME model to home care services will be performed continuously at local project group meetings during the RCT in the intervention municipalities.

The intervention with TIME is based on a theoretical framework of complexity science and a biopsychosocial understanding of crises [34, 35]. In this framework, frailty means that the frail patient is prone to instability caused by complex interactions among biological, psychological and social stressors [14, 35]. If this instability rises, it eventually culminates in the development of a crisis that, according to our description of crises, demands immediate and new actions. Describing frailty and crises as complex phenomena sets the stage for why interventions should be constructed and implemented as flexible complex interventions to be able to assess, prevent, and resolve crises [24]. This also includes the choice of design and methods for testing the effectiveness of an intervention.

The primary purpose of this study is to test the effectiveness of an adapted version of a biopsychosocial person-centred model (TIME) to prevent and resolve crises for frail community-dwelling people receiving home care services. We hypothesise that the adapted version of the TIME intervention would prevent and resolve crises compared to a control condition consisting of usual care.

This is a pragmatic study (effectiveness study) focusing on testing the effect of an intervention under real-world conditions. We will use a participatory action research (PAR) design in a cluster randomised controlled trial (RCT) with two parallel groups: intervention municipalities (IMs) and control municipalities (CMs) [23, 36]. Figure 1 shows a flow chart of the clusters and individuals throughout the phases of the trial based on the power calculation.

PAR aims to ensure adaptation of the components of the intervention to the local context [23, 36]. This will be done by establishing a local project team in each municipality consisting of two representatives from the research group and local participants (representatives of local managers, staff, and GPs). These groups will adapt the implementation process and the TIME model according to the cyclic Deming process: plan, do, check, act, and adjust [37]. The implementation process will be adapted to the local context to fit with the organisational structure already established in home care services. For example, the content and time used for educational and training purposes can vary depending on the educational level and established educational arenas. These variations will be registered during the process. For the TIME model, the core components (functional) of the intervention (TIME) must be fixed, and other components can be adapted to the local context and the patients included. An assessment phase must be performed, but what to assess and the types of clinical scales to be used can vary (processes). Case conferences using the inductive cognitive ABC model for the analyses of crises must be conducted, but the timeframe and participants of these conferences will vary. The action and evaluation phase must be conducted, but the timeframe and types of actions and evaluations can vary. These adaptations of the model will be accurately mapped in each setting during the study, and this mapping will be a part of the process evaluation study in the PRACTIC study [29]. In this trial, we will follow the recommendations from the CONSORT statement for randomised trials of nonpharmacologic treatments [38].

This trial is part of the larger PRACTIC (Preventing and Approaching Crises for Frail Community-dwelling Patients Through Innovative Care) study. The trial will include approximately 30 randomly selected municipalities and their home care services from all health care regions in Norway. From each region, a sample of small, medium, and large municipalities will be invited to participate. From these 30 included municipalities, 150 users of home care services and their next of kin will be invited to participate in the trial.

The inclusion criteria for patients are as follows: (1) in need of home care services, (2) a score ≥ 5 on the Clinical Frailty Scale (indicating mild to severe frailty) [39], and (3) perceived by the home care service as being in an unstable situation with a high risk for acute institutionalisation or showing resistance to care. The only exclusion criterion is an expected short life expectancy (i.e. < 4 weeks).

Approximately four project nurses from each home care service unit in the municipality will be trained to obtain informed consent. For more details of the recruitment of these project nurses and the training session, see the “Intervention description {11a}” section.

N/A. Participant data and biological samples are not collected in ancillary studies.

Participation in the project is voluntary. The participant can withdraw consent at any time and without giving any reason. This will have no consequences for further home care services. We have not established any other criteria for modifying the allocated intervention for any given trial participant.

Members of the research team at the AFS research centre will contact the local project nurses in each municipality twice with 2-month intervals. A structured telephone interview with a checklist that consists of the core components of TIME (i.e. the assessment phase, case conferences, the actions taken and systematic evaluations of these actions) will be carried out.

The services and treatment measures provided are generally recognised care and treatment. The patients in the control group will receive care and treatment as usual. We have no other criteria for concomitant care permitted or prohibited during the trial.

Not applicable, as there is no post trial care.

The study timeline for enrolment, intervention, and assessment is described in Table 3. In addition, description of the questionnaires and the time points at which they are administered is provided in Table 4.

The local project nurses (See: Who will take informed consent? {26a}) will support the research team in the recruitment of participants. Care providers for the intervention will, however, be the permanent staff in the home service units. Each municipality must recruit approximately five patients, and the project nurses have one week to include participants in the study. These nurses will know the patient well and assess whether the inclusion criteria are fulfilled for each patient for participation in the study. The project nurses will ask eligible patients if they wish to participate when they visit the patient in their home. Each IM and CM will be reimbursed 900 EUR for the time spent in this recruitment process.

This trial will be conducted as a cluster randomised controlled study, where each municipality represents a cluster. The clusters will be randomised to either receive the locally adapted TIME intervention (the intervention group) or care and treatment as usual (the control group). The intervention is a biopsychosocial intervention that involves the entire interdisciplinary team and staff in the home care service units of the participating municipalities to optimise the approach towards a group of patients in the municipalities. Thus, without cluster randomisation, the study runs the risk of implementing all or parts of the intervention model among individual control patients in the same municipalities [52].

The municipalities will be stratified by size into three blocks to ensure approximately the same number of patients in the two trial arms. These blocks are (1) small municipalities, (2) medium-sized municipalities, and (3) large municipalities. The municipalities within each block will be randomly assigned to either the intervention group or the control group. A statistician independent of the project management team and the municipalities will perform a computer-generated randomisation procedure.

The randomisation results for the municipalities will be sent from the independent statistician by email to a researcher (AV) in the research team by an encrypted file and a connection key. Only this researcher and the research team will know the randomisation results.

A statistician independent of the project management team and the municipalities will perform the randomisation by the use of a computer procedure. The project management team will provide the home care services in the municipalities with the randomisation and allocation results immediately after the randomisation procedure. The intervention will start with the educational sessions (11a) within 1 to 2 weeks after randomisation.

The baseline assessments, before randomisation, will be performed by the same data assessors by visiting the participants in their homes and interviewing them, the next of kin, and the staff members who know the patient best. The data assessors in the study will be blinded to whether the municipality is allocated to the intervention group or the control group at the assessments at 3 months and 6 months. These data assessments will be performed by telephone by interviewing the staff 3 months and 6 months after baseline. The statistical data analyst will also be blinded to the allocation results. After baseline and the allocation process, the home care services, the training group, and the patients cannot be blinded.

Unblinding during the study will not be permitted, and the data assessors will be blinded to the randomisation result.

The data assessors are specially trained nurses from the project’s research centre who are not affiliated with the municipalities. All data assessors are registered nurses with substantial experience and formal training in the use of the assessment scales, and they will attend a 1-day course about the use of the assessment scales before the start of the trial. A description of the questionnaires, including data assessors, respondents, and the time points at which they are administered, is provided in Table 4.

The participants will be assessed at three time points, at baseline and at 3 and 6 months, during the 6-month RCT. The participants in the intervention group will also be followed up by phone by the local project nurses using a checklist for adherence to the study protocol 2 and 4 months from baseline. For all participants who withdraw from the study before the 6-month follow-up, the date and reason for discontinuation will be registered by the local project nurses and reported to the research team.

The data protection official, Innlandet Hospital Trust, has approved the applications for data entry and security. The collected data from the sampled patients will be deidentified and stored on a secured research server at Innlandet Hospital Trust.

Specially trained nurses (data assessors) from the project’s research centre who are not affiliated with the municipalities will travel to the municipality to assess baseline data using a paper case report form (CRF). Data will also be assessed by the same data assessors at 3 and 6 months using the CRF, such as telephone interviews with nurses from the municipality. The data assessors will deliver the CRFs directly to the researchers in the project group at the research centre when they have been completed. The data will then be scanned and transferred into SPSS (Statistical Package for the Social Sciences [53]) at the project’s research centre. The SPSS file will be stored on a secure server at Innlandet Hospital Trust. The data will be kept for 5 years after the end of the project for control reasons.

N/A. Collection, laboratory evaluation, and storage of biological samples for genetic or molecular analysis is not applicable in this study.

N/A. No preliminary analyses will be performed, and we have not developed any stopping guidelines. All treatment measures developed and implemented for the patients in the intervention group are regular and known clinically accepted procedures and care, with no new experimental measures or treatments.

N/A. There are currently no planned additional analyses.

To handle protocol nonadherence, the analysis will be performed as an intention-to-treat analysis. The linear mixed model handles unbalanced data by allowing the inclusion of all available information, including dropouts and missing data.

The last version of the full protocol and the datasets will be made available on the project’s website: www.​practic.​no. Participant-level data and the statistical code will only be accessible to the central research team.

This study is a part of the larger PRACTIC study. The project is owned by the Research Centre for Age-Related Functional Decline and Disease (AFS), Innlandet Hospital Trust, Ottestad, Norway. The core research group with the project leader, work package leaders, postdocs, PhD students, and a project coordinator will have day-to-day responsibility for the project. A project group comprising the core research group and other research staff will have regular meetings once a month to discuss the project. The PhD students and research staff will collaborate with participating municipalities in the project and have regular dialogue. The PhD students are responsible for coordinating and collaborating with the training group for educational sessions during the intervention and the data assessor group throughout the RCT.

The project has organised two reference groups, one with end-users (patients and next of kin) and one with staff from home care services, including GPs. The reference groups will meet with the central project group regularly during the project period. The design of the project will enable close cooperation with end-users and stakeholders in the municipalities through the establishment of local project groups in each municipality.

This trial does not require periodic inspections of cumulative outcome data. The study involves little risk for the participants, so there is no need for periodic inspection of accumulating outcome data by a formal committee such as a data monitoring committee (DMC).

The data protection official, Innlandet Hospital Trust, has approved applications for the trial. If participants have questions about privacy in the project, they can contact the institution’s data protection officer. If any adverse events occur, this will be reported to the data protection officer.

In this study, no new or experimental treatments will be provided to the participants, and they will receive only traditional health care and assistance for their needs. The intervention mainly comprises an introduction of a more person-centred way of working for the staff that is very unlikely to cause harm to the patients. The principal investigator (PI) for the study is responsible for the daily operation of the study, and the participants or their next of kin can contact the project leader in any cases of harm or questions pertaining the study.

N/A. There is no requirement for the frequency of or procedures for auditing the conduct of the trial.

In the event of important protocol changes, a change notification will be sent to the data protection officer, Innlandet Hospital Trust.

The datasets supporting the conclusions of this article will be available on the website for the PRACTIC study. Three papers will be published in international peer-reviewed open access journals. In addition, the results from this study will be presented orally at national and international congresses. The results of the project will also be communicated to participants in the train the trainer course in TIME.

Due to the multimorbidity and diversity of the patients receiving home care services, care interventions can hardly be introduced as standardised solutions based on single diagnoses but should be implemented through holistic approaches embracing multimorbidity and different functional impairments. The PRACTIC study will enhance innovation in the development of new knowledge and the development of a new biopsychosocial approach towards each patient. This process will be adapted to local structural conditions. The project is likely to enforce systematic cooperation between home care services and GPs, where diagnostic work-up and follow-up will be one of the most important tasks. For home care services, this means a cultural change from a mainly task-oriented service based on support for the activities of daily life to an interdisciplinary assessment and follow-up of functional impairment and diseases for these patients. If successful, because of the trial’s pragmatic character, the intervention can easily be implemented in health care services with minimal extra resources. Improving the approaches to crises may also reduce the use of specialist health care services.