Improving mental health and end-of-life care in residents of nursing homes: A cluster randomized clinical trial of efficacy

Related to this successful trial, we published/re-submitted/submitted further 19 COSMOS related contributions in 2017. COSMOS data are not only used by PhD candidates directly related to the trial, Irene Aasmul and Christine Gulla who will complete their PhD in 2018. Data are also used by two research-line students (Torstein Habiger, Tony Elvegaard), other PhD candidates related to the Center for Elderly and Nursing Home Medicine (SEFAS) (Kjersti Blytt, Ane Erdal), two master degree students (Lillian Birkeland, NHH; Ingunn Fleten Mo, physiotherapy), two medical students (bachelor, UiB) (Jørgen Petersen, Ingrid Aurlien), and three PhD students from abroad (Marjan Laekeman, University of Bamberg, Germany; Charlotte Griffeon and Paulien van Dam, both University of Leiden, The Netherlands). The main COSMOS article has been just submitted and is under review: Husebo BS, Ballard C, Aarsland D, Selbaek G, Slettebo DD, Gulla C, Aasmul I, Habiger T, Elvegaard T, Testad I, Flo E. A 9-month complex intervention of communication, pain assessment and treatment, medication review and activities versus control to improve the quality of life in nursing home patients (COSMOS): a cluster-randomized controlled trial. In brief, the methods is describes as a cluster-randomized, single-blind, controlled trial including patients with and without dementia (N=723; 65 years) from 33 nursing homes. We randomized 67 units (clusters) to care as usual or intervention groups in which healthcare staff received standardized education and on-site training over the course of 4 months and follow-up at month 9. Primary outcome was change in QoL (QUALIDEM, QUALID and EQ-VAS) from baseline to month 4. Secondary outcomes were clinical impression of change (CGIC), total medication, activities of daily living (ADL) and staff distress. We also analyzed the efficacy of the intervention between baseline and month 9. Interestingly, we found that the QoL was reduced in the intervention group compared to control, measured by QUALID (mean difference in change 1·3, 95% CI 0·1 to 2·6, p=0·04) and QUALIDEM (-2·4, -3·8 to -0·9, p=0·002), at month 4. However, follow-up analysis from month 4 to 9 showed an intervention effect for EQ-VAS (p=0·003) and QUALIDEM total score (p=0·01; Care relationship p=0·02, Positive affect p=0·04, and Social relations p=0·01). Moreover, significant improvements were demonstrated for some secondary outcomes assessed from baseline to month 4 (CGIG, reduction of medication, agitation, depression, and staff distress). We concluded that although the QoL temporarily decreased during the active COSMOS intervention, QoL reversed significantly during follow-up compared to control. Patients randomized to intervention received less medication, and staff reported less distress by COSMOS and judged the intervention as clinical relevant change. Furthermore, significant reduction of agitation and depression could be provided by the trial. The non-pharmacological intervention in our trial provided an education programme with a train-the-trainer approach and close follow-up of nursing home staff during the intervention period. Results are of key importance for clinicians and stakeholders because the combination of repeated communication, pain assessment and treatment, medication review and activities may change everyday practice in a real-world nursing home setting. Noticeably, these results were achieved in the frailest population with a mean age of 87, and with severe degree of dementia, multimorbidity and polypharmacy. The gravity of these patients? condition was demonstrated by the fact that one in five patients died during the trial. We are currently involved in developing a mobile application, which will contain relevant research results from the COSMOS-project and a description of the implementation procedures. The format limits the themes to be presented in a concrete and easy-to understand manner. The application is intended to be equally useful for nursing home staff, patients, their relatives and the public. The application will be lounged February 2018. Our COSMOS experiences are also baseline for a new RCN financed study LIVE@Home.Path for home-dwelling people with dementia and is the prerequisite for an ongoing application to the Horizon 2020 call: Novel patient-centred approaches for survivorship, palliation and/or end-of-life care (SC1-BHC-23-2018). The PALLIDEM trial will focus on PALLiative care Innovation for Community Dwelling people with DEMentia. Together with researchers from nine European countries, Australia and Japan, we will use our COSMOS experiences to scale up research in primary care, dementia care, palliative care, and innovation, in Europe.

Nursing home (NH) patients have complex mental health problems, disabilities and social needs, compounded by widespread prescription of psychotropic drugs. To preserve their dignity and quality of life (HQoL) is an important goal of our society. This can only be achieved within NHs that offer high competent conditions of treatment and care. KOSMOS (Kommunikasjon, Omsorg i livets slutt, Smertebehandling, Medikamentvurdering, Occupational therapy, og Sikkerhet) is a practical intervention aimed to improve c linical and psychiatric challenges in NH patients. The KOSMOS intervention combines most effective research results to improve staff competence, and patients mental health, safety, HQoL, and to reduce psychotropic drug use and costs. The efficacy testing of KOSMOS includes systematic literature review, a 4 months pilot study, 1.5-year RCT, interviews with relatives, and a dissemination plan. Quantitative and qualitative measures will be collect at baseline, months 6, 12, 18, and after death (interviews). Development of manuals and education of trained NH therapists are planned. 380 patients, 65 and older, will be recruited from 20 NHs in Bergen, Stavanger, and Oslo. NHs will be randomized to either KOSMOS or current best practice. Outcome measures: CMAI, DEMQOL; NPI-NH; ADL; Cornell; MOBID-2; QUIS; psychotropic use; drug-related problems; pain and symptom management; fall-diary; cost-utility analysis (RUD-FOCA); hospital admission; and mortality. Statistical analyses include characteristics between 2 grou ps (Chi square, Mann-Whitney U), ANCOVA, ICC, p-values for each time-point and outcomes. Collaboration between researchers at the Universities of Bergen, Oslo, and Stavanger is established. Colleagues from the COST-Action TD1005, Karolinska University, St ockholm and Kings College, London are engaged in this RCT. They all worked active with the project plan, which apply for a project leader (50%) and 2 PhD candidates (100%).