Table 3 Continuity of care

Case management

Ellis 2010

Patients with stroke

I: Stroke liason worker

Subjective health, function, participation, death, institutionalisation, mood, stroke related knowledge, health service utilisation, patient satisfaction

Patients with mild to moderate disability had a significant reduction in dependence (OR 0.62, 95% CI 0.44 to 0.87) and there were reports of higher patient and carer satisfaction however there was no other evidence that the intervention improves outcome.

Design: SR

C: Alternative care or no post-discharge care

Size: 16 studies

Setting: Predominantly community

AMSTAR: 8/11

Chesnut 1999

Patients with TBI

I: Long term care coordination

General functional status

There was a lack of high quality studies and studies included in the review reported conflicting results therefore the authors were unable to make clear recommendations on the evidence for this approach.

Design: SR

Size: 3 studies

Setting:

Community

AMSTAR: 3/11

Allen 2009

Patients with stroke

I: 6 months of follow up contact from an Advanced Practice nurse who worked with the GP to implement a care plan, organised further services and provided education

NIHSS, TUG, physical performance test, mortality, institutionalisation, QOL, mgt of post-stroke complications, stroke knowledge and lifestyle modification

There was little difference between groups at 6 months. The intervention group had slightly improved lifestyle management and stroke knowledge.

Design: RCT

Size: N = 380

Setting: Acute to community

United States

PEDro: 8/10

C: Usual care plus provision of written stroke related education.

Bell 2005

Patients with TBI

I: Regular phone calls over 9 months upon discharge to follow-up any issues, identify concerns and provide information, mentoring, goal setting, reassurance and referral to community resources.

Function, community integration, neurobehavioural functioning inventory, Glasgow outcome scale, QOL, emotional state

The intervention group had significantly better scores on functional status and perceived quality of wellbeing than the control group however the magnitude of the effect is unclear as the outcome was a composite measure.

Design: RCT

Size: N = 171

Setting: Rehabilitation to community

United States

PEDro: 8/10

C: Usual care

Bell 2008

Patients with mild TBI

I: Regular phone calls over 3 months post injury after presentation to the ED, a contact phone number and additional information about brain injury and where to get help

Head injury symptoms, QOL. PHQ, role performance, community participation

Patients in the intervention group reported fewer symptoms 6 months post injury than the control group (6.6 difference in adjusted mean symptom score, 95% CI 2.2 to 5.2)

Design: RCT

Size: N = 366

Setting: ED to community

United States

PEDro: 8/10

C: Usual care (patient handout and outpatient treatment if prescribed).

Bell 2011

Patients with TBI

I: Regular phone calls for up to 21 months post injury from a case manager. The purpose of the calls was to help participants to identify, prioritise and solve problems as independently as possible

Function, level of disability, participation, symptoms, QOL, vocational status

No significant differences were found between groups for any of the measures at either 1 or 2 years post injury.

Design: RCT

Size: N = 433

Setting: Rehabilitation to community

United States

PEDro: 6/10

C: Usual care

Trexler 2010

Patients with ABI

I: Allocated to ‘resource facilitators’ who contacted participants every 2 weeks (via telephone or home/community visits). A large focus of the facilitator was returning the patient to work (CM)

Participation, self reported health

Levels of participation improved more in the intervention group (F = 9.11) and more of the intervention group were employed at the time of 6 month follow up (n = 7 vs n = 4).

Design: RCT

Size: N = 22

Setting: Community

United States

PEDro: 4/10

C: Usual care (no ‘resource facilitator’

Early supported discharge

Fearon 2012

Patients with acute stroke

I: Early supported discharge

Death or long term dependency, length of stay, ADL function, subjective health status, mood, carer outcome, patient and carer satisfaction

Appropriately resourced Early Supported Discharge models can reduce length of stay (equivalent to approximately 7 days). Patients receiving ESD are more likely to be independent and living at home OR 0.80 (95% CI 0.67 to 0.97)

Design: SR

C: Other models of care

Size: 14 studies

Setting: Acute stroke to home

AMSTAR: 9/11

Short term package

Andersen 2000

Patients with stroke

Ia: Three home visits from a geriatric rehabilitation physician for medication review, referral/liason with other services, information

Function, ADLs, mortality, institutionalisation and readmission

Significantly less hospital readmissions in both intervention groups compared to the control group at 6 months (26% vs 34% and 44%).

There was no difference in functional outcome between groups at 6 months.

Design: RCT (three arms)

Size: N = 155

Setting: Rehabilitation to community

Ib: Home visits (average of 3) from the hospital physiotherapists for instruction and education

Denmark

PEDro: 7/10

C: Standard aftercare (may have included outpatient rehabilitation and home care)

Forster 2009

Patients with stroke

I: Follow up assessment from a stroke nurse 5–6 months after stroke. Issues identified in the assessment were managed in a standardised manner

Function, mood, satisfaction with care, caregiver burden

There were no real differences between groups at 12 month follow up however patients reported improved satisfaction with care in some areas.

Design: RCT

Size: N = 265

Setting:

Community

United Kingdom

C: Letter sent to GP recommending 6 month review

PEDro: 8/10

Ghaffar 2006

Patients with mild TBI

I: Patients were followed up in a multidisciplinary TBI clinic within 1 week of injury and offered pharmacotherapy, PT, OT and supportive psychotherapy if required

Symptoms, General Health questionnaire, cognition

In general there were no significant differences between group however a small subgroup (those with a premorbid psychiatric history) appeared to benefit from treatment, reporting lower levels of depression at 6 months (F = 6.8).

Design: RCT

Size: N = 191

Setting: ED to community

Canada

PEDro: 5/10

C: Usual care (no follow up arranged)

Wade 1997

Patients with head injury

I: Routine follow up from an occupational therapist or psychologist 7–10 days post injury with organisation of further follow up as required

Symptoms

As a whole, there were no significant differences between groups however subgroup analyses revealed that patients in the control group with a more severe head injury were more likely to have continuing problems at 6 months.

Design: RCT

Size: N = 1156

Setting: ED to community

C: Usual care (no routine followup)

United Kingdom

PEDro: 3/10

Wade 1998

Patients with head injury

I: Routine follow up from an occupational therapist or psychologist 7–10 days post injury with organisation of further follow up as required

Symptoms

The intervention group reported fewer or less severe concussion symptoms and less disruption of social activities at 6 months than the control group.

Design: RCT

Size: N = 314

Setting: ED, hospital and community

C: Usual care (no routine followup)

United Kingdom

PEDro: 5/10

Ytterberg 2000

Patients with stroke

I: All day follow up visit one month after discharge

Self reported health status

There were no significant differences between groups

Design: RCT

C: Usual care with no specific follow-up arranged

Size: N = 111

Setting: Community

Sweden

PEDro: 2/10