Early ed Discharge trialists (supported by Stroke Association and CHSS) Facilitating early supported discharge after stroke Peter Langhorne Professor of Stroke Care University of Glasgow Craig Anderson (Auckland) Erik Bautz-Holter (Oslo) Paola Day (Manchester) Martin Dennis (Secretariat) Jean Douglas (Administrator) Bent Indredavik (Trondheim) Peter Langhorne (Coordinator) Nancy Mayo (Montreal) Gordon Murray (Statistician) Michael Power (Belfast) Helen Rodgers (Newcastle) Ole Morten Ronning (Akershus) Sally Rubenach (Adelaide) Anthony Rudd (London) Nijasri Suwanwela (Bangkok) Gillian Taylor (Statistician) Lotta Widen-Holmqvist (Stockholm) Charles Wolfe (London) ESD trialists. Cochrane Library 2005; Langhorne et al, Lancet 2005 Early supported discharge Summary of the early supported discharge story so far What can we conclude Is ESD effective? How to implement ESD? Can we explain effects? Remaining uncertainties Background and definitions ESD services aim to accelerate discharge home from hospital and provide rehabilitation / support in the home setting 1
Integrated stroke service - objectives Primary prevention? Integrated stroke service - objectives Primary prevention? Patient Patient Outpatient Assessment Prevention Inpatient Assessment Acute care Prevention ilitation Outpatient Assessment Prevention Early ed Discharge service Inpatient Assessment Acute care Prevention ilitation Continuing support & prevention Continuing support & prevention Long term support and re-assessment of needs Long term support and re-assessment of needs Conventional services Early supported discharge Admission Discharge Review Admission Discharge Review Acute ilitation Acute Home Home 2
Early ed Discharge Potential risks and benefits? Potential benefits Home better setting for rehabilitation Favoured by patients and carers Free hospital beds Reduce costs Early ed Discharge Potential risks and benefits? Potential risks Unable to manage medical problems Strain on patients and carers Expensive if done well Increase costs Why we need randomised trials Participant Matched groups R No contamination Treatment Control Little loss of participants Unbiased follow up Blinded outcome assessment 12 randomised trials (1659 patients) 3
Main outcomes Patient selection in ESD trials Primary Death or dependency (Rankin 3-5) at end of scheduled follow up (6; 3-12 months) Secondary Death Death or institutional care Length of hospital stay ADL, extended ADL score, Patient and carer satisfaction Condition unstable Severe dependency Confusion Live in PNH Patients selected Medically stable Persisting disability Able to comply Live locally 40 (12-70)% of admissions Good recovery Conventional services Early supported discharge Admission Discharge Review Admission Discharge Review Acute ilitation Acute Home Home 4
Early supported discharge Multidisciplinary team (MDT) coordination/ delivery Early supported discharge Multidisciplinary team (MDT) coordination Admission Discharge Review Admission Discharge Review Acute Acute Home Home ESD team: PT, OT, nursing etc ESD team Community rehabilitation services Early supported discharge No multidisciplinary team (MDT) coordination Early supported discharge vs Conventional care Death or dependency Admission Discharge Review Acute Home Various community rehabilitation services 5
Early supported discharge vs Conventional care Death or dependency Early supported discharge vs Conventional care Death or dependency Absolute outcomes at 6-12 months post stroke (additional events for every 100 patients treated) Secondary outcomes Resource outcomes Outcome Alive Living at home Independent in ADL Additional benefit (95% confidence interval) 0 (-3, 2) 5 (1, 9) 5 (0, 9) Significance P=0.76 P=0.02 P=0.04 ESD services, compared with conventional services, resulted in: Reduced length of hospital stay by 8 days (5-11; P<0.0001) No increase in readmission to hospital over subsequent 6-12 months (27% vs 25%; P>0.1) 6
Secondary outcomes ADL, EADL, mood subjective health Secondary outcomes Patient and carer satisfaction Patient outcomes Improved EADL scores (SMD 0.12; P=0.05) No significant differences in ADL, subjective health or mood scores Carer outcomes No significant difference in subjective health status or mood scores Outcome Patients satisfied with outpatient care Carers satisfied with outpatient care ESD service 69% 82% Control service 61% 74% Difference per 100 treated (95% CI) 9 (2, 17) 7 (-2, 17) P P=0.02 P=0.14 Economics of ESD services Individual trial analyses (London, Newcastle, Adelaide, Stockholm, Trondheim) indicate; Reduction in hospital bed use Increase in community costs Overall there were modest savings with ESD services Costs and benefit vary with severity of stroke? Early supported discharge services Promising role for the future Not applicable to all stroke patients (50%?) Can accelerate discharge home Appear to improve subsequent longer term recovery (remaining independent at home) Probably resource neutral (in theory) ESD trialists. Cochrane Library 2005; Langhorne et al, Lancet 2005 7
Implementation of results? Early ed Discharge team Organisation, staffing and skills of the ESD team Community or hospital-based team Setting of the services Stroke unit or general wards Urban versus rural communities Patient selection Carer present Stroke severity Pathway of care Service subgroups Death or dependency Service subgroups Death or dependency Service subgroups Death or dependency NB Greater reduction in length of stay with hospital outreach Recent trial in rural setting 8
Implementation ESD teams Coordinated MDT input appears to be important Physiotherapy, occupational therapy, nursing (SALT and medical input) or community-based? Access to services and equipment Access to stroke unit / inpatient services Most trials of stroke rehabilitation or neurological rehabilitation teams Patient selection Death or dependency Patient selection Death or dependency Implementation - patients Clinical benefit seen in patients with mildmoderate initial stroke severity Barthel index at least 10/20 in first week post stroke However most reduction in length of stay seen in severe subgroup 28 day reduction versus 4 day reduction Evidence of ESD benefit seen only on mild-moderate severity group 9
Implementation care pathway Explanation of results? Admission to hospital Discharge from hospital Discharge from ESD Contact with patient/carer Identify key worker Home assessment Plan discharge Agree rehabilitation goals Implement rehabilitation plan Access relevant services MDT review of progress Negotiate withdrawal admission discharge ESD discharge Key components Contact with key worker Home assessment Plan discharge Agree rehabilitation goals Therapy input and access relevant services MDT review of progress Planned withdrawal ing evidence Systematic review (?) No RCTs Systematic review (?)? 2 Systematic reviews Indirect evidence Carer training? admission discharge ESD discharge Explanation of results? Key components Contact with key worker Home assessment Plan discharge Agree rehabilitation goals Therapy input and access relevant services MDT review of progress Planned withdrawal ing evidence Systematic review (?) No RCTs Systematic review (?)? 2 Systematic reviews Indirect evidence Carer training? admission discharge ESD discharge Explanation of results? Key components Contact with key worker Home assessment Plan discharge Agree rehabilitation goals Therapy input and access relevant services MDT review of progress Planned withdrawal ing evidence Systematic review (?) No RCTs Systematic review (?)? 2 Systematic reviews Indirect evidence Carer training? 10
Therapy input at home The Outpatient Service Trialists. Lancet 2004. Regular therapy input (PT, OT or mixed) increased activities of daily living: EADL score SMD 0.17 (0.04-0.30) Walker MF et al. Stroke 2004. Community occupational therapy significantly improved personal and extended activities of daily living. Better outcomes were found with targeted interventions Carer training? (Kalra et al. BMJ 2004) Trained care givers experienced less caregiving burden, anxiety or depression and had a higher quality of life. Patients reported less anxiety and depression and better quality of life in the caregiver training group Patients' mortality, institutionalisation, and disability were not influenced by caregiver training. Remaining uncertainties Optimal implementation of ESD teams Specialist vs generic rehabilitation team outreach vs community inreach Urban vs rural settings Remaining uncertainties Patient selection Clinical / economic impact in different patient groups Coordination with other services Day hospital, home therapy, stroke liaison worker Monitoring implementation 11
Early supported discharge services Promising role for the future Not applicable to all stroke patients (<50%) Can accelerate discharge home Appear to improve subsequent longer term recovery Best results with ESD services coordinated and provided by a multidisciplinary rehabilitation team and targeted at mild-moderate stroke patients Several uncertainties remain Key 1) TIA 2) Mild 3) Moderate 4) Severe Integrated stroke service - components Rapid access neurovascular clinic (1) Patient (1) (2,3,4) (2) Comprehensive stroke unit (2,3) (3,4) Early supported discharge Continuing rehabilitation Inpatient rehabilitation Long term support and re-assessment of needs Thank you 12
Implementation - examples outreach team Stockholm (Stroke 1998;29:591-597) Trondheim (Stroke 2000;31:2989-2994) Community inreach team Newcastle (Clin 1997;11:280-287) London (BMJ 1997;315:1039-1044) Adelaide (Stroke 2000;31:1032-1037) Why we need systematic reviews (meta-analyses) Trials of small size Single centre Publication bias Lack power and prone to chance Limited external validity (applicability) Need to examine all similar trials 13
Remaining uncertainties Optimal implementation of ESD teams Specialist vs generic rehabilitation team outreach vs community inreach Urban vs rural settings Patient selection Clinical / economic impact in different patient groups Coordination with other services Day hospital, home therapy, stroke liaison worker Monitoring implementation Why we need randomised trials In order to be confident about our estimates of benefit and harm we need to look at studies with: Adequate randomisation matched patient groups Blinded (masked) follow up unbiased assessment of outcomes Complete follow up unbiased assessment of outcomes 14