Frailty and multimorbidity is one of the biggest challenges of todays health care due to the demographic development with more and more elderly surviving many diseases that was mortal just a few decades ago. Health care in Sweden is one of the best in the world, yet many frail older people do not receive appropriate health care. Several reports have described a fragmented care with lack of good quality due to lack of a holistic view on the patient and her situation. At the same time the care given is costly and ineffective with a lack of continuity. Additionally, there are a lack of hospital beds in Sweden due to the ageing population and a cut down of beds in the last decades.

There is though a method, Comprehensive Geriatric Assessment (CGA), where studies have shown many benefits for older patients including less need of hospital beds, better functional outcomes and a better chance of living at home after an Acute Hospital Discharge, but this method is not much used in Sweden.

The aim of this study is to show that CGA in an Out-patient Care Setting save hospital beds without decreasing the quality of care measured by sustained functional capacity. Other parameters to be evaluated are mortality, degree of frailty, health economy, quality of life, and cognition.

Frailty and multimorbidity is one of the biggest challenges of todays health care due to the demographic development with more and more elderly surviving many diseases that was mortal just a few decades ago. Health care in Sweden is one of the best in the world, yet many frail older people do not receive appropriate health care. Several reports have described a fragmented care with lack of good quality due to lack of a holistic view on the patient and her situation. At the same time the care given is costly and ineffective with a lack of continuity. Additionally, there are a lack of hospital beds in Sweden due to the ageing population and a cut down of beds in the last decades.

There is though a method, Comprehensive Geriatric Assessment (CGA), where studies have shown many benefits for older patients including less need of hospital beds, better functional outcomes and a better chance of living at home after an Acute Hospital Discharge, but this method is not much used in Sweden.

The aim of this study is to show that CGA in an Out-patient Care Setting save hospital beds without decreasing the quality of care measured by sustained functional capacity. Other parameters to be evaluated are mortality, degree of frailty, health economy, quality of life, and cognition.

Survey of the field Frailty has become a central concept in elderly research in recent years. A common definition is a state of decreased reserve resistance to stressors as a result of cumulative decline across multiple physiological systems, causing vulnerability to different outcomes, such as falls, hospitalization, institutionalization and mortality. The prevalence of frailty increases with age and frail older people are at risk of further deterioration if their needs are not acknowledged. The Swedish National Board of Health and Welfare has given priority to the development and use of targeted needs assessment for the frail older people. Assessments that are both comprehensive and person-centred are required to achieve satisfactory and appropriate care to older people with complex needs.

Frailty is a dynamic condition, conceptualized as a continuum from not frail, to pre-frail, and frail. Those who are considered as pre-frail are more likely than non-frail to develop full frailty if no specific interventions are implemented. Appropriate nutrition and exercise are suggested as first-line treatments for frailty but other treatments should be considered depending on the individual causes of frailty. The goal is to identify needs and provide support to help older people to be as independent in their daily living as possible.

Previous research has found that interventions including CGA results in functional benefits for frail older patients including a better chance of living at home after an acute hospital care-period. CGA involves a multidimensional team approach assessing medical, functional, psychosocial and environmental needs . Several studies have implied less need of hospital beds because of fewer readmissions when care is given according to CGA. As there are a substantial lack of hospital beds in Sweden, which results in patients placed in corridors, hazardous discharges and problems with the working environment, it would be important to find an intervention that could both save hospital beds and give a better care for these patients.

As frailty can be addressed by care based on CGA it is important to detect frailty, but also to avoid the use of a complex CGA when not needed. A recent review have shown that frail elderly benefits more of CGA than a more mixed population of elderly. There is unfortunately no easy and overall accepted instrument to detect frailty. In research the criteria developed by Fried have been most frequently used - but is resource-demanding and seldom used outside academic research. Another more easy accessible method have been developed by Rockwood and have been chosen from the The International Consortium for Health Outcomes Measurement (ICHOM) as the preferred instrument to detect and measure frailty. The instrument is not well known in clinical practise in Sweden, why it could be interesting to test whether it applicable in Swedish Health Care.

There have been only very few studies on CGA in Out-patient Care - but one recent randomized controlled study conducted in Sweden 2011-2013 showed a nearly 30% reduction of need of days in hospital and a similar reduction of mortality three years after randomization. The investigators would therefore perform a similar study in a slightly different population based on experience from the former Swedish study in order to see whether the results are reproducible in another health care setting in Sweden.

Methods This project is a single-centre trial with the base in Helsingborg Hospital.

Participants who meet the inclusion criteria are identified by a special register in the ER able to detect when a patient has visited the ER several times within a specified period. The register is managed by an administrator without any coupling to the research project. The ER administrator will every month send a list of possible participants to a research nurse. She will control the list for exclusion-criteria (living in nursing home or outside the nearest municipalities to Helsingborg Hospital).

All remaining possible participants will receive a letter of invitation explaining the purpose of the study and the study protocol. Within seven days from the sending of the letter the subjects are contacted by telephone for verbal informed consent and booking an appointment at their home for further information, written consent and baseline assessment by a study nurse. If an old person due to cognitive decline is unable to answer the study questions or give informed consent, a proxy is contacted. Those accepting to participate will be randomized by computer-generated numbers into two study arms: intervention or control group. This information will be kept as a randomisation master list with the powered number of participants (450) by a research administrator. When written informed consent has been given and the baseline assessment done, the visiting study nurse will contact the administrator who will give the patient a randomisation number consecutively according to the master list. A letter will be sent within a week to inform the patient about to which group he or she has been allocated.

The intervention: After baseline assessment the intervention group receive an interdisciplinary assessment based on CGA, tailored to each participant for 24 months. The first visit is either at the hospital or in the patient's home, depending on the participant's condition and ability to travel. The multidisciplinary teams consisting of physician, nurse, pharmacist, physiotherapist and occupational therapist with access to more competences if needed. The team works according to CGA, and has the primary and continuing responsibility for the patient during the study period of 24 months. The CGA is person-centered and unique for each patient, but is based on the core elements such as socio-demographic background, social network, health and medical history, medications, functional status, cognitive status, nutritional status, somatic status and psychosocial status including depression, providing a comprehensive assessment tailored for each person. Structured questionnaires and validated measures will be used.

The team will be able to give a proactive care based on the CGA in collaboration with the primary care and the municipality. The relatives will be involved when possible. The intensity of follow-up depends on the participants' needs.

Usual care (control group) Participants allocated to the control group receive health care from their general practitioner, the community services and in- and outpatient hospital care.

Both the intervention and control group have access to the primary care centers, the hospital and the ambulatories on equal conditions.

Follow ups of outcome measures will be performed in the older person's home at 12 and 24 months after inclusion.

Sample size A power calculation has been made, based on the primary outcome variable, number of days in hospital. Based on a former similar study in a slightly younger population there will be an assumed difference between the intervention group and the control group of 4,1 days in hospital in 24 months (11,1 days in the intervention group and 15,2 in the control group) with standard deviation of 15 days in both groups. To be able to detect a difference between the intervention and the control groups with a two-sided test and with a significance level of α=0.05 and 80% power, at least 211 participants in each group will be needed. There will be almost no loss to follow-up of of this outcome based on previous research of this patient-group and intervention, so a total of 450 persons will be included.

Statistical analysis:

Both descriptive and analytic statistics will be used, in order to compare groups and for analyses of changes over time. Non-parametric statistics will be used when ordinal data are analysed. Otherwise, parametric statistics will be used. Besides descriptive statistics, the chi2 and Fisher's 2-tailed exact test will be used to test differences in the proportions between the groups, and odds ratio (OR) will be used to compare outcomes between groups. A value of p < 0.05 (2-sided) will be considered significant. The analysis will be made on the basis of the intention-to-treat principle. Given the old age of the participants, a relatively high drop-out rate will be inevitable, and missing data will not be at random. Simply analyzing complete cases is not relevant and might lead to bias. Therefore, the approach of data imputation will be the replacement of missing values with a value based on the median change of deterioration between baseline and follow-up of all who participated at follow-up. The reasons for this imputation method is 1) the study sample (frail older people) is expected to deteriorate over time as a natural course of the aging process and 2) reasons for not fulfilling the follow-ups are often deteriorated health. Worst-case change will be applied for those who have died before follow-up.

An interim-analysis will be made on the primary outcome and mortality after one year of an independent researcher.

Time schedule The planning of the study will be completed during early autumn 2016. Recruitment of clinical team-members are ongoing and in October 2016 the Out-Patient Care Team will start to work based on CGA to be ready to include patients.

November 2016 - December 17 Inclusion-period. A preliminary study has revealed 50 - 60 eligible patients per months in Helsingborg. With an attrition rate of 30-40% there will be around 35 patients to be randomized a month. To reach the empowered number and with lower inclusion rate during summer and vacations it will take 12-14 months to include all participants and the study period for each participant will be 2 years.

January 17 - March 19 Follow up of outcome measures. August 19 All data collection completed. Quality assurance of data will be done by the project assistant and the project leader.

August 19 - December 19: Analysis and article writing (researchers), as well as presentation at conferences (for professionals as well as researcher; local, national and international, done by professionals and researchers).

Inclusion Criteria:

• 75 Years
• 3 visits in the Emergency Room (ER) during the previous 18 months
• 3 diagnoses according to ICD-10.

Exclusion Criteria:

• people living outside the Municipalities of Helsingborg, Åstorp, Bjuv and Höganäs municipality
• living in nursing homes