Effect of stroke early supported discharge on length of hospital stay: analysis from a national stroke registry

OBJECTIVE: The first observational study to investigate the impact of early supported discharge (ESD) on length of hospital stay in real-world conditions. DESIGN: Using historical prospective Sentinel Stroke National Audit Programme (SSNAP) data (1 January 2013-31 December 2016) and multilevel modelling, cross-sectional (2015-2016; 30 791 patients nested within 55 hospitals) and repeated cross-sectional (2013-2014 vs 2015-2016; 49 266 patients nested within 41 hospitals) analyses were undertaken. SETTING: Hospitals were sampled across a large geographical area of England covering the West and East Midlands, the East of England and the North of England. PARTICIPANTS: Stroke patients whose data were entered into the SSNAP database by hospital teams. INTERVENTIONS: Receiving ESD along the patient care pathway. PRIMARY AND SECONDARY OUTCOME MEASURES: Length of hospital stay. RESULTS: When adjusted for important case-mix variables, patients who received ESD on their stroke care pathway spent longer in hospital, compared with those who did not receive ESD. The percentage increase was 15.8% (95% CI 12.3% to 19.4%) for the 2015-2016 cross-sectional analysis and 18.8% (95% CI 13.9% to 24.0%) for the 2013-2014 versus 2015-2016 repeated cross-sectional analysis. On average, the increased length of hospital stay was approximately 1 day. CONCLUSIONS: This study has shown that by comparing ESD and non-ESD patient groups matched for important patient characteristics, receiving ESD resulted in a 1-day increase in length of hospital stay. The large reduction in length of hospital stay overall, since original trials were conducted, may explain why a reduction was not observed. The longer term benefits of accessing ESD need to be investigated further. TRIAL REGISTRATION NUMBER: http://www.isrctn.com/ISRCTN15568163.