A follow-up audit was undertaken following ASU implementation over the period 1 October 2008 to 30 September 2009 (2009 audit). The 2006 audit focused on key processes which were thought to be problematic at the time such as aspirin administration, multidisciplinary team involvement in stroke patient care and inpatient waiting times before transfer to a rehabilitation service. Multidisciplinary input was provided by one of 3 teams involving medical, physiotherapy, speech language therapy, social work and nursing staff. Patients were cared for by one of six general physicians.

We were also interested to examine the effectiveness of a stroke unit in a medium-sized New Zealand hospital as previous studies were undertaken in larger hospitals.Methods The initial audit was a retrospective case note based study of all strokes presenting to Hutt Valley Hospital between January 1 and Decem(2006 audit) when there was no organised stroke care at the hospital. Stroke care at Hutt Valley Hospital was provided by general physicians in undifferentiated medical beds until an ASU was created in October 2008.This study is a comparison of audits, the first undertaken prior to ASU implementation and the follow up performed during its first year of operation to see if improvements had been made to processes and outcomes of stroke care. Taylor et al15subsequently found that the introduction of a clinical care pathway for stroke had no impact on outcomes.Hutt Valley Hospital is a 260-bed hospital located in the lower North Island servicing a population of 140,000. He suggested a clinical pathway to improve stroke care and ideally an ASU, but felt that implementing an ASU in a smaller hospital may not be practical. Only 2% of patients had a swallow assessment documented in the first 24 hours of admission. He found serious deficiencies in stroke clinical assessment and process of care. McNaughton14 audited 50 consecutive discharges with a diagnosis of stroke using the Royal College of Physicians Stroke Audit tool. This was not a before and after study.Stroke care has previously been studied at Hutt Valley Hospital. 94% of patients with stroke were cared for on the Stroke Unit, and guideline adherence was generally high.

Both these studies suggested significant improvement in process but no clear improvement in outcomes.Gommans et al13 audited a random selection of 50 patients admitted with stroke to Hawke's Bay Hospital's Stroke Unit. Mortality decreased from 14.0 to 8.8%, but only 33% of patients were cared for in the ASU, and stroke unit subgroup mortality was worse probably due to the admission of more unwell patients. Significant improvements in process were noted, particularly in the case of swallow assessments and allied health input. Total length of stay (Pre 12.0 days, Post 15.7 days, p=0.14) and 30 day mortality (Pre 24% vs Post 25% p=0.94) were not different.Di Matteo et al12 audited outcomes following the creation of a 12 bed ASU at Middlemore Hospital by retrospectively reviewing the case notes of a random sample of patients admitted pre (1999 n=100) and post (2001 n=100) ASU implementation. Improvements were noted in process of care. Hanger et al11 used the Royal College of Physicians (London) stroke audit tool to examine cohorts of 119 patients admitted before and 72 patients admitted after the institution of a 15 bed ASU at Christchurch Hospital. The components of stroke unit care that influence outcome are still unknown and some researchers have labelled it a 'Black Box'.8 There is some evidence that process of care in stroke units is important in improving outcomes.9,10There have been three previous New Zealand studies looking at stroke unit care.

Others had some form of organised stroke service.4Research into discrete interventions has shown that prescription of aspirin,5 intravenous thrombolysis for selected patients6 and decompressive craniectomy7 are important in reducing mortality and dependency in acute stroke. These provided service for 48% of the New Zealand population. Over 7000 people suffer from stroke each year and less than half are alive and independent at 1 year.¹ A Cochrane review in 2006 showed that co-ordinated stroke care in dedicated stroke units reduced mortality and dependency.2, Current New Zealand guidelines for stroke management advocate for the establishment of stroke units in all hospitals.3In 2007 15% of New Zealand hospitals had an acute stroke unit (ASU). Stroke is a leading cause of death and disability in New Zealand.

Thomson j hudson