Gothenburg Very Early Supported Discharge study (GOTVED): a randomised controlled trial investigating anxiety and overall disability in the first year after stroke

The Gothenburg Very Early Supported Discharge study (GOTVED) is a randomised controlled trial with blinded assessors. This trial is registered on clinicaltrials.gov (identifier: NCT01622205). Participants were enrolled in the trial from September 2011 to April 2016.

All patients admitted to one of the stroke care units at Sahlgrenska University Hospital of Gothenburg were consecutively screened (Fig. 1). Patients fulfilling the inclusion criteria were informed by a research coordinator about the study and asked if they wanted to participate. Written informed consent was obtained from the participants or from their closest relative. A block randomization model was applied (20 patients in each block; 10 in the VESD group and 10 in the control group). A person not otherwise involved in the study performed the allocation. This was done by putting a paper slip with group information in envelopes, then sealing, mixing and numbering them. The research coordinator opened an envelope after inclusion, and then the blinded assessor and the nurse in the stroke team were informed of the inclusion. The blinded assessor worked at a different ward at the hospital to minimise learning of the allocation by chance. The CONSORT checklist was followed. The stroke subtypes were confirmed by imaging, and treatment of thrombolysis or thrombectomy was recorded.

Inclusion criteria were confirmed stroke according to World Health Organization criteria [19, 20], age > 18 years, residence within 30 min by car of the stroke unit, a National Institute of Health Stroke Scale (NIHSS) [21] score of 0–16 points, which corresponds to mild to moderate stroke [20], a Barthel Index (BI) [22] score of > 50 points on day two [23], and a Montreal Cognitive Assessment (MoCA) [24] index of < 26 if BI = 100. Thus, exclusion criteria were a NIHSS score of > 16 and a BI score < 50. Patients with a life expectancy of < 1 year (e.g. with severe malignancy) or who were unable to speak or communicate in Swedish prior to stroke were also excluded.

The primary outcome as stated in the protocol was anxiety [25]. The anxiety was assessed with the Hospital Anxiety and Depression Scale-Anxiety subscale (HADS-A) [26], which was administered 5 days (± 1 day), 3 months (± 3 days), and 12 months (± 1 week) post-stroke. HADS is a 14-item self-assessment scale that can be divided into two equal parts, HADS-A for anxiety (score 0–21) and HADS-D for depression (score 0–21). A higher score indicates more symptoms of anxiety or depression [27]. The secondary outcome was the patients’ degree of overall disability, measured by the modified Rankin Scale (mRS) [28]. The scale runs from 0 to 6, with zero indicating perfect health without symptoms and six indicating death. The scores can be dichotomised, and a score of ≤2 indicates functional independence [29]. ADL function was assessed by the BI, which ranges from 0 to 100, with higher scores indicating a higher level of ADL independence [30]. Cognitive functions were assessed with the Montreal Cognitive Assessment (MoCA) [24] with a score of ≥26 indicating normal cognitive functioning [31]. Stroke-related neurological deficits were assessed with the National Institute of Stroke Scale (NIHSS) [21] by the nurse at the stroke unit during the first 2 days; the value at day two was used for inclusion. BI and MoCA were administered by occupational therapists 36–48 h after arrival at the stroke unit. A trained, blinded researcher not working at the stroke unit performed the HADS-A and mRS assessments at baseline and all assessments at three and 12 months post-stroke. Exact time points for all assessments are listed in Table 1.

Prior to discharge of the patients in the VESD group, as a part of the person-centred intervention, a goal-setting meeting was held at which the patient was asked to formulate his or her goals based on the Canadian Occupational Performance Measure [32]. These goals guided the focus of the rehabilitation. Examples of goals were to be able to go to the local store to buy milk, to be able to hang the laundry, or to be able to travel on the tram to a daughter, or to manage the bills. The intervention had a person-centred approach based on the person, her or his context, history, next of kin, individual strengths and weaknesses and expressed personal goals [33].

A rehabilitation team made up of physiotherapists, occupational therapists, and a stroke nurse from the stroke care unit continued the rehabilitation in the patient’s home. At discharge, the intervention group received an individual schedule for the first week of home rehabilitation. The intervention comprised 2–4 visits per week by the physiotherapist and/or occupational therapist and 1–2 visits by the stroke nurse. The intervention could include varied activities or methods to think about different ways of adapting in difficult situations. For some patients, the intervention was to try the intended activities with safe support so that they could feel secure in their performance. The patient and the VESD team together decided when the support from the team should end; the maximum length was 4 weeks after discharge. In connection with the discharge from the VESD, if necessary, the patients were referred to outpatient rehabilitation that would carry on the rehabilitation afterwards. Information and support to next of kin and the home care service on how to best support the patient to reach the decided goals was also of importance [25].

Patients in the control group were discharged according to the department’s usual routine. They had no goal-setting meeting and were not followed up by a multidisciplinary team from the stroke unit. However when needed, they were referred to continued rehabilitation or/and support (e.g. outpatient rehabilitation with a physiotherapist and/or an occupational therapist, home care service).

A power calculation was performed based on the level of anxiety (assessed with the HADS) [25]. With a power of 80% and a p-value of 0.05 (2-sided test), a sample size of 44 patients per group were needed to detect a 4 point difference [34, 35]. As deaths may occur or consent be withdrawn, we aimed at 55 patients per group. During the inclusion process, more participants than expected dropped out. Therefore, after 2 years of screening, we decided to include a total of 140 patients. Intention to treat analyses were performed, and for missing observations, the “last value carried forward” was used meaning that everyone who started the study was included in the analysis, even though they did not complete the treatment. This meant that dropouts were also included in the analysis. The HADS outcome is presented both as a continuous variable and as a trichotomized. Shifts in the proportions of anxiety disorders were analysed for the intervention group and control group at admittance and after three and 12 months and reported using bar graphs. For this analysis the HADS-A scores at baseline and after three and 12 months were trichotomized according to the subscale scores proposed in the assessment tool: “no annoying anxiety” (0–7) = 1, “mild to moderate anxiety” [8‐10]=2, and “occurrence of anxiety disorders” (> 10) = 3 [26]. This was also performed with the mRS: 0 = no symptom at all, 1 = no significant disability despite symptoms, 2 = slight disability, 3 = moderate disability, 4 = moderately severe disability. For each outcome, the common odds ratio and 95% confidence intervals were reported for the shift in the direction of a better outcome in both groups. The chi-square test and Mann-Whitney U-test were used to test for group differences in descriptive data, and to evaluate whether there was any statistically significant difference in outcome between those who were included in the study and those who were not. Descriptive statistics are expressed in percentages, mean ± SD or median and interquartile range (IQR), as appropriate. The Wilcoxon signed rank test was used to assess the change in proportions between admittance, 3 months and 12 months after onset with the HADS-A and the mRS. The effect size was reported according to Cohen [36]. A two-sided value of p ≤ 0.05 was considered statistically significant. Statistical analyses were performed using IBM SPSS statistics for Windows, version 24.0 (IBM Corp., Armonk, NY, USA).