Procedure Participants were recruited consecutively from the rehabilitation unit Vikaergaarden in the period from January 17, 2012 to May 29, 2015, and from the rehabilitation unit Thorsgaarden from October 20, 2014 to May 29, 2015. The random allocation of the participants to an intervention group (IG) or a control group (CG) was done by an independent external organization ("TrialPartner", Public Health and Quality Improvement, Central Denmark Region). The permuted block sizes stratified the randomization according to sex, age, and place of referral. The randomization took place within three days after the participants' arrival to the rehabilitation unit. Owing to the nature of this study, it was impossible to blind participants and their relatives to the allocation group. The project manager was blinded to the study outcomes, which were collected from the registers or by the blinded research occupational therapist. Rehabilitation units' staff, particularly physiotherapists, were not blinded.
Care in the CG The patients were referred for rehabilitation either from hospital or home by the hospital personnel or by the home care staff.The typical standard rehabilitation and care program lasts five weeks. The interdisciplinary approach is based on the patient's whole situation, capability and wishes/needs. On the first day of rehabilitation, the patient's functional status is observed by the rehabilitation unit's physiotherapists and occupational therapists, and a nutritional screening is performed by the rehabilitation unit's nutritionist.The team members discuss the patient's discharge destination and necessary arrangements with the patient and his/her relatives at the mid-term meeting and before discharge from the rehabilitation unit. Municipality nurse participates in these meeting personally or by telephone. Destination after discharge is based upon the patient's motivation, functional and medical status.
The patient's GPs visit the patients during the stay if required or occasionally by own initiative depending on practice routine and geographical distance. GPs mostly visit frail and high-risk elderly patients especially if recently hospitalized. Acute medical aid is called for in case of illness after 4.p.m. and on weekends and public holidays.
Care in the IG Participants in the IG had the same access to usual care and additionally underwent a CGC performed by a physician specialized in geriatric medicine. The intervention included medical history, physical examination, blood tests, medication adjustment and treatments, including intravenous antibiotics or blood transfusions conducted by the geriatrician at the rehabilitation units. The geriatrician was employed 18.5 hours per week and was present at the rehabilitation units for four days a week, and could be contacted on telephone for any reason by participants, their relatives, or the units' staff on weekdays from 8 a.m. to 3 p.m.
BaselineAssessments Before randomization baseline characteristics were registered from medical records by the project manager, comprising age, gender, place of referral, marital status, residence, previous diagnoses, list of medications and comorbidity burden by Charlson Comorbidity Index (CCI). Baseline functional status and quality of life measures were assessed on day 3 from admission to the rehabilitation units by a research occupational therapist.
Statistics The numbers of hospital admissions, ED visits, ambulatory contacts, and GP contacts will be compared by Incidence Rate Ratios (IRR) with 95% confidence intervals (CI). IRR will be calculated using negative binominal regression with adjustment for mortality by including the risk time as an exposure variable.
Mortality rates after 30 and 90 days will be calculated as percentages of deaths in total population per group. In non-survivors and in case of missing baseline observations, the missing measures will be set to the worst possible values. The worst value imputation method will be used in all other cases of missing values.
The mean group values will be analyzed using repeated measurements mixed model. Dichotomous variables will be created to investigate distribution between participants who worsened/did not change versus participants who improved their functional status or quality of life, and analyzed using logistic regression. Results will be expressed as odds ratios (ORs) with 95% CIs.
Subgroup analyses will be conducted to explore if effects of intervention when compared to usual care depended on the baseline comorbidity burden low/moderate (CCI 0-2) versus high (CCI ≥3 points). The subgroup analyses will be carried out for the functional status and quality of life.
There were no drop-outs during the study and data on healthcare utilization are complete. Intention-to-treat analyses will be performed in the all 368 participants. The two-sided significance level of 5% will be used for evaluation of statistical significance in the primary and secondary endpoints.
(The materials and methods are described in detail elsewhere: Zintchouk D, Lauritzen T, Damsgaard EM. "Comprehensive Geriatric Care Versus Standard Care For Elderly Referred To A Rehabilitation Unit - A Randomized Controlled Trial". J Aging Res Clin Practice 2017;inpress. Published online December 15, 2016, http://dx.doi.org/10.14283/jarcp.2016.126).
