Embed Size (px)
Citation preview
SCAP: Stroke Clinical Audit ProcessAssessing and addressing unwarranted clinical variation
Thursday 28th April 2016. Reducing Unwarranted Clinical Variation in Stroke
Conjoint Associate Professor John Worthington| Liverpool Health Service and Ingham Institute UNSW
The ACI Stroke Network and ACI have taken BHI’s UCV
data to the bed-side in search of local solutions to
unwarranted clinical variation
*The Insights Series: 30-day mortality following hospitalisation, five clinical
conditions, NSW, July 2009 – 2012. Analysis methods are on the BHI web-
site in the ‘Spotlight’ document.
After the ACI pilot audits the BHI analysis was modified to measure
outcomes by hospital of first presentation for this hospital identified
analysis. Variation is measured against an arithmetic mean. Smaller
hospitals were excluded due to small numbers and wide confidence limits
The identified BHI 30 day mortality data on 5 conditions was
released 6-8 months earlier than expected and before planned
meetings with hospital managers and clinicians.
Why? The BHI publication of 30 day ischaemic stroke mortality 2009-2012,
with identification of hospitals. Published December 2013.*
NSW hospitals look after 11,000 strokes of all types per year
A minority of hospitals provide organised stroke care.
23 Acute Thrombolysis Centres (ATCs) are now nested in 36 acute stroke units across
NSW. Nine other hospitals have stroke services. New stroke units, a further stroke
service and three new ATCs are coming on line as result of local efforts and SCAP.
There are 186 sites in NSW with some ED role delineation, 79 with level 3-6 role
delineation. Forty nine hospitals see more than 50 strokes of all types a year, 33
hospitals see more than 100 strokes a year and 7 see more than 400 pa.
The 30 and 365 day ischaemic stroke mortality in NSW is 17 and 27%, respectively
(Gattellari et al, Cerebrovascular Diseases, 2011).
Where stroke units are implemented in NSW there has been a 30% improvement in
mortality and in discharge destination (Gattellari et al, Stroke 2009.)
NSW outcomes for stroke compare favourably with OECD countries and other states
(BHI 2012 and 2013), however, they report unwarranted clinical variation between sites.
3
Ischaemic stroke care in NSW
4
Causes of death after stroke
A hectic three weeks
Management in the first 2-3
weeks has a major impact
on mortality, long-term
function and discharge
destination
Determinants of poor outcomes
•Aspiration, sepsis and fever
•Venous thrombosis
•Hypoxia
•Dehydration
•Tachycardia eg: poor AF rate control
Stroke requires close attention from an experienced multidisciplinary team
in a stroke unit of co-localised beds over days and weeks
5
48
20
46
5
19
0
10
20
30
40
50
%
*Retrospective medical record audit of 5,413 stroke patients in acute NSW public hospitals throughout 2000-2014. Median age 78 years (Q1: 68, Q3: 84), 51% male and 93% with ischaemic stroke.
Eight percent experienced a severe complication while in acute hospital care.Purvis T, Longworth M, Kilkenny M, Worthington J, Pollack M, Levi C, Cadilhac D
Common severe complications in hospital shown as a percentage of all documented complications
* Includes aspiration pneumonia and other chest infection
Stroke progression results from raised
intracranial pressure, dehydration, other
metabolic disturbance and sepsis and is
relatively low in well organised stroke
care. It can reflect quality of care
ACI Audit: Proportion of stroke complications in NSW 2000-14*
6 pilot sites: Comparison of processes
expected to influence stroke patient outcomes
Unwarranted clinical variation in stroke is explicable variation. At present stroke patients do not
always receive evidenced-based care. This may be the result of being admitted to a smaller
hospital with no organised stroke care and little prospect of providing it, admission to a
hospital where stroke unit care could reasonably be provided but no unit has been established,
because patients fail to reach stroke unit beds in a hospital with a stroke unit or because of
variations in the quality of care delivered in existing stroke units.
Pilot results (and
methods) provided a
proof of concept for
the SCAP project
There is substantial evidence around what constitutes good ischaemic
stroke care.
Major elements of good stroke care include:
Stroke units. With co-localised stroke beds served by a
multidisciplinary stroke team that uses evidenced-based pathways
improve stroke outcomes by approximately 30%, at all ages, in
NSW.1 All are eligible for Stroke Unit care. New NWAU adjuster.
Clot-busting. IV rt-PA within three hours, reduces death and disability
by 44% (Cochrane), with more modest benefits at 3-4.5 hours
(favourable Odds Ratio 1.34).2,3 There is an all-hours cost-of-
readiness and no DRG. Eligibility around 16% of all strokes in high
performance settings. New IV Thrombolysis code in July.1Gattellari et al Stroke 2009; 40: 10-7.
2 Wardlaw et al, Cochrane Database of Systematic Reviews. 2003 (3).3.Emberson et al. Stroke Thrombolysis Trialists’ Collaborative Group. Lancet 2014, Published online.
Evidence based practice in ischaemic stroke
DISCHARGE DESTINATION Home Nursing home Death Other*
10 NON-PRINCIPAL REFERRAL HOSPITALS (METRO) Age > 85 years
Before ASU 20.3% 12.9% 26.8% 40.0%
After ASU 28.7% 10.3% 19.7% 41.4%
10 NON-PRINCIPAL REFERRAL HOSPITALS (METRO) All adults
Before ASU 38.7% 6.3% 13.8% 41.2%
After ASU 44.5% 4.9% 10.5% 40.2%
*transfer to other hospitals/change in type
Outcomes for ischaemic stroke before and after introduction of
stroke units in 10 Non-Principal Referral NSW hospitals
p<0.001 (significant main effect and interaction type*time). Controlling for: age, co-morbidity (modified Charlson Index), sex, marital status, country of birth, hours on mechanical ventilation, insurance status, and clustering of outcomes by hospital in GEE multivariate model. Gattellari et al Stroke, 2008.
968975
56
Pre-programme Post-programme
Stroke units improve the quality of stroke care
• Clinical care plan is defined as evidence of a written plan by health professionals to avoid complications.
• **Stroke clinical pathway is defined as a structured tool detailing the activities of stroke care during hospital admission.
Site enhancement led to
improvements in clinical care
processes
• Within 24 hrs of admission
– 7% more patients
received brain imaging
– 19% more patients
were swallow-tested
• Clinical care plans were
written for an additional
27% of patients
• Stroke clinical pathways
were recorded for an
additional 49% of patients
Patient undergoes clinical processes
within 24 hours of admission13
% of patients
72
45
65
16
Post-programmePre-programme
Clinical care plan and stroke clinical pathway
developed during admission13
% of patients Clinical care plan*
Stroke clinical pathway**
Swallow-tested
Brain imaging
(CT & MRI)
Modified after Cadilhac DA et al. Qual Saf Health Care. 2008.
Enhancement and clinical process adherence
10
Adherence to nominated clinical process of care indicators for the six hospitals that
participated in the Rural Stroke Project and Stroke Clinical Audit Process
Not every improvement was
maintained or reached
acceptable levels
Stroke care and complications in NSW*
11
0.1 1 10
Age
Independent prior
Imparied speech*
Unable to walk*
Arm deficit*
Incontinent at 72 hours
Haemorraghic stroke
Team meeting
Care plan
No severe complication Severe complication
Factors associated with
severe complications**
Stroke Pathway
Severe complicationN = 448
No severe complication
N = 4,965p value
Patient Characteristics
Male 209 (47%) 2,503 (51%) 0.1
Age median (Q1, Q3) 81 (74, 86) 77 (67, 84) <0.001
Independent prior^ 256 (61%) 3,438 (72%) <0.001
Stroke type/severity at presentation
Haemorrhagic stroke 372 (85%) 4,466 (94%) <0.001
Impaired speech 338 (82%) 3,074 (65%) <0.001
Arm deficit 370 (86%) 3,368 (70%) <0.001
Unable to walk 321 (80%) 2,536 (58%) <0.001
Incontinence at 72 hours
341 (79%) 1,835 (40%) <0.001
Hospital factors
Rural location 259 (58%) 2,884 (58%) 0.9
Neurologist 101 (23%) 1,296 (26%) 0.1
Bedside factors
Stroke unit care 136 (30%) 1,770 (36%) 0.03
Brain scan within 24 hrs
384 (86%) 4,288 (88%) 0.5
Physiotherapy within 24 hrs
92 (21%) 1,271 (26%) 0.02
Regular neurological observations
303 (69%) 3,185 (65%) 0.1
Team meeting 97 (22%) 833 (17%) <0.01
Stroke pathway 115 (26%) 1,694 (35%) <0.001
Aspirin within 24hrs# 150 (42%) 2,627 (60%) <0.001
**Results of bivariable analyses
*Retrospective medical record audit of 5,413 stroke patients in acute NSW public hospitals throughout 2000-
2014. Median age 78 years (Q1: 68, Q3: 84), 51% male and 93% with ischaemic stroke. Eight percent experienced a
severe complication while in acute hospital care.Purvis T, Longworth M, Kilkenny M, Worthington J, Pollack
M, Levi C, Cadilhac D
ACI stroke audits were
carried out pre- and
post-stroke unit
implementation and in a
wide range of
metropolitan and rural
hospitals over almost 15
years.
12
Improving ischaemic stroke outcomes in NSW
The potential years of life lost due to all stroke types has fallen by 16% over 10
years in NSW which is midrange among other OECD countries
In 2011 the age standardised 30 day
mortality of ischaemic and haemorrhage
stroke in those over age 45 years was
11.5 and 29.6%, having fallen by 19 and
13%, respectively, over the 10 years
(2003-2013).
ACI has funded the Stroke Clinical Variation Statewide Strategy (SCVSS) & Stroke Clinical
Audit Program (SCAP) July 2014-December 2015
Stroke Clinical Audit Process (SCAP). Unwarranted Clinical Variation Taskforce approved
extension to the successful Pilot. Completion of thirty supervised audits over two years.
SCAP. Detailed auditing and feedback of thirty hospitals. Measuring adherence with processes
expected to impact on stroke patient outcomes, benchmarked against earlier and other site
audits with analyses by Florey/NSRI and the SCAP team
Validation of routinely collected data used in Bureau of Health Information analyses through a
data linkage with SCAP audit data. In scope with BHI.
Reducing the burden of audit and prospective data entry with development of a Stroke bundles
of care for Electronic Medical Record (Build C) with requests to provide data extraction tools.
Home to Outcome Study (H2O). The OHMR funded and ACI partnered Ingham Home to
Outcome (H20) study. A pioneering data-linkage to better describe and measure the whole
stroke journey across NSW hospitals. Uses 10 data bases containing ambulance, NSW ED
and hospital admission, rehabilitation, death and readmission data.
13
ACI actions: Examining clinical variation to improve stroke care.
The NSW Stroke Network and ACI are providing local clinicians and
managers with the data and analyses needed to improve stroke care
SCAP audits: Unwarranted clinical variation
Example of hospital feedback•Median age 74 yrs. 7% intracranial haemorrhages, no IV
‘lysis, 1 palliative care and 2 deaths.
•AF identified 21%, diabetes 25%, high cholesterol 31%, a
previous stroke or TIA in 23 and 10%.
•Risk factors on medication at admission AF 75%, Diabetes
84%, high cholesterol 58%, IHD 64% and HT 77%. Only
56% with previous stroke or TIA were on antithrombotics.
•2 transfers in. None with a t/f protocol. Average time 1.5
hours; 100% presented at transferring hospital by
ambulance. No in-patient strokes.
•Direct to Stroke Unit/ICU/HDU/CCU 85%. 67% SU.
•Neuro obs 24hrs 68%, 96% brain imaging <24 hrs, 3%
Clinical Care plan, 99% a d/c strategy, 13% Stroke Pathway
and none had an MDT Family meeting.
•Echo and Duplex (36 and 76%). MRI 61%. 88% ‘unable to
walk’ had heparin/LMW heparin; 100% (4) NBM at 48 hours
received IV/NG fluids.
•64% received aspirin < 24 hours and 72% with IS were on
an antithrombotic at discharge. New statin 44%.
•Speech pathologist in 24 hours 67% (86% if speech
impaired). 31% documented swallow<4 hrs.
14
Ix Hospital 9 2003, 2005, 2007 and 2015
Ix ‘Hospital 6’ 2006, 2009 and 2014
Example: Hospital 6 Pilot Audit Results 2013
•Rural SU and ATC. Similar results to 2008/9
•55% transferred in (one for rehab). Hub and spoke!
•Average age 71 years
•35% had AF
•15% a previous stroke
•All were admitted to the stroke unit!
•75% were on a stroke clinical pathway during the
admission.*
•65% had a CT within 2 hours and 100% in 24
hours.
•Stroke investigation rates shown in figure
•100% received neurological observations in the
first 24 hours
•72% received aspirin in the first 24 hours.
•Documented swallow assessment in 4 hours of
40% (45% in speech impaired)*
15
No hospital unit performed
consistently well across all clinical
care processes that are likely to
influence patient outcomes. Where
outcomes appeared worse the gaps in
evidence-based care were generally
greater
*There was local surprise at rates of pathway use and swallow assessment with an immediate QI response
Assume Nothing!
Clinical variation: Measuring and improving care.
SCAP and pilot audit, analysis and feedback
Adherence with bed-side processes known to improve
patient outcomes and experience
Access to desired investigations
Use of a stroke clinical pathway
Access to stroke unit beds
Access to a multidisciplinary team
Evidence-based prescribing
Prevention and timely treatment of stroke
complications
16
SCAP audits: Unwarranted clinical variation
No hospital unit performed uniformly well across all processes.
Brain imaging in the first 24 hours varied between 46% and 100%.
Cardiac echocardiography 0 to over 90%. Carotid duplex 0-86%.
Stroke pathway use varied between 0 to over 90%. Two major
teaching hospitals do not use a pathway which has been shown to
reduce complications
Two major hospitals with higher than expected BHI 30 day mortality
estimates admitted only 50 and 60% of their ischaemic stroke
patients to stroke unit beds.
Highest rate of VTE prophylaxis in patients with difficulty walking
was 88%, only fourteen sites exceeded 50%. Five sites, including
two stroke units had rates lower than 15%.
17
SCAP audits: Unwarranted clinical variation
Some hospitals identified as Acute Thrombolysis Centres to which
ambulances were being directed only provided ‘clot-busting’
treatment to 1-2% and others exceeded 20% in the audit samples.
Enhancement from state-wide programmes or local initiatives are
seen to improve overall adherence with desirable processes.
Adherence did not always reach appropriate levels even where
services were enhanced eg VTE prophylaxis.
Discharge on antithrombotic in IS varied widely from 46 to 93%
Nursing and allied health determined processes and timely access
to allied health usually improved or were maintained over time
Several processes dependent on medical decisions eg: prescribing
were not maintained or did not reach acceptable levels.
18
19
Unwarranted clinical variation and clinical process adherence
Unenhanced and enhanced sitesTargeted efforts to enhance sites results in better
adherence with processes which are expected to
improve outcomes
There is a relationship between
adherence with processes expected
to improve outcomes and BHI
estimates of 30 d mortality
Hospitals are ranked from left to right by increasing mortality estimate
SCAP audits: Average rates of investigation across
Unenhanced and Rural and Metro Enhanced sites
20
Hospital 1: Investigations over 4 audits
The rates of investigation were lower at
unenhanced hospitals some of which had
no onsite CT scanning, with an average
of 74% receiving brain imaging within 24
hours. CT rates at two Unenhanced sites
were 36 and 43%. Documented carotid
imaging and echocardiography rates
were zero at some sites
21
SCAP: Process measures at 8 Unenhanced Rural sites N=495
0
20
40
60
80
100
120
Hospital 16 Hospital 19 Hospital 21 Hospital 22 Hospital 25 Hospital 27* Hospital 28* Hospital 29*
Process measures: SCAP audits of 8 unenhanced rural sites SU/HDU/ICU
24hr NeuroObs
Stroke Clinical Pathway
Swallow<24
DC Antithrombotics
Asprin<24hrs
DC on Statin
VTE Proph.
Linear (24hr NeuroObs)
Linear (Swallow<24)
Linear (Asprin<24hrs)
Linear (VTE Proph.)
Hospitals ranked from left to right with increasing mortality estimates
Hospital 16 has an 18% 30 day IS mortality and risk-standardised mortality ratio of 1.27
22
SCAP: Process measures at 9 Enhanced rural sites N=510
0
20
40
60
80
100
120
Hospital 2 Hospital 3 Hospital 6 Hospital10
Hospital17
Hospital18
Hospital20
Hospital23
Hospital26
%
Hospitals ranked from right to left by increasing estimated mortality
Process measures: 9 Enhanced Rural sitesBrain Imaging*
Physio*
Speech*
OT*
Documented Swallow*
Documented Swallow**
MDT Family Meeting
Any SU/HDU
Stroke pathway
Clinical care plan
Linear (Any SU/HDU)
Linear (Stroke pathway)
23
SCAP: Process measures at 12 Metropolitan hospital sites
N=784
0
20
40
60
80
100
120
%
Process measures: SCAP audits of 12 metropolitan hospitals SU/HDU/CCU
24hr Neuro Obs
Stroke pathway
Swallow Test<4hrs
%Discharged on Antithrombotics
Aspirin<24hrs
DC on new Statin
%VTE P'laxis if immobile
Linear (SU/HDU/CCU)
Linear (Stroke pathway)Hospitals ranked from left to right by increasing estimated mortality
24
Hospital,Audit No,
(Mean age),No. strokes pa 2013-14,
ACI audit periods.
BHI 30d Ischaemic Mortality and
RSMR* 2009-12. (Crude SCAP audit
mortality all strokes)
SU/HDU/ICU
Bed(%)
24 hr
Neuro Ob's (%)
Stroke
ClinicalP’way (%)
Swallow test< 4 hrs
(%)
%D’chargedon
A’thrombotic
Aspirin at 24 hours
(% IS)
Pall’
Care(N)
% D/C on
Statin
%VTE
P’laxis
(Not mobile)
1. NPR SU N=68
N=159
12/11-2/2013
Awaiting re-audit data
9% and 0.57(0%)
91 96 97 45 78 63 0 75 50
4.PR N=20
N=240
8/11-11/11
11% and 0.78 100 95 45 70 84 58 3 63 0
5.PR N=80
N=503
3/12-7/12
12% and 0.84 89 94 0 10 93 56 3 53 58
7.NPR N=40
(76yrs)
N=81
2005 and 3/13-3/14/15
10% and 0.89(10%)
85 3 65 55 75 59 2 44 35
8. NPR N=80
(70yrs)
N=296
2003, 05, 07 and ‘15
12% and 098
(1.5%)
77 91 79 20 74 70 0 42 25
9. PR N=79
(75 yrs)
N=457
2001,2007+1/14-4/2015
13% and 0.99(13%)
86 89 63 20 77 70 3 75 37
10. PR N=79
(71 yrs)
N=477
2003, 05, 07+9/13-4/14
13% and 1.06 90 95 70 37 78 58 1 51 71
11.NPR N=79
(74 yrs)
N=266
2003, 2005,2007, Jan 13-
Dec14
14% and 1.1(3%)
85 68 13 31 74 64 1 44 88
Less RED numbers and more GREEN squares are better
25
Tabulation of process measures expected to influence
outcomes at 29 SCAP audit sites
Ranked by BHI 30 day mortality risk
(or crude audited rate if not available)
*Risk Standardised Mortality Ratio. BHI Insight
series.
**Note: a calculated audit sample mortality of 0%
reflects a lack of access to the files of deceased
patients at time of audit.
NA= Not available
Red numbers indicates low measures. Blue indicates
higher measures.
Tile shades of Green, Yellow and White indicate the
ranking of measured processes within columns.
Palliative care in Grey is not ranked.
More GREEN squares and
less RED numbers are better
Hospital,Audit No,
(Mean age),
No. strokes pa 2013-14,ACI audit periods.
BHI 30d Ischaemic Mortality
and RSMR* 2009-12. (Crude
SCAP audit mortality all
strokes)
SU/HDU/ICU
Bed
(%)
24 hr
Neuro
Ob's (%)
Stroke
Clinical
P’way (%)
Swallow test< 4 hrs (%) %D’charged
on A’thrombotic
Aspirin at 24
hours (% IS)
Pall’
Care
(N)
% D/C on Statin
Hospital 1. NPR SU N=68
N=159
12/11-2/2013
Awaiting re-audit data
91 96 97 45 78 63 0 75 50
Hospital 2. N=40
(78 yrs)
N=79
2011+4/13-8/14
77 80 85 87 72 68 1 68 79
Hospital 3. 14 N=70
(79 yrs)
N=180
2006,2009+2/13-1/14
96 100 100 44 79 52 0 52 61
Hospital 4. N=20
N=240
8/11-11/11
100 95 45 70 84 58 3 63 0
Hospital 5. N=80
N=503
3/12-7/12
89 94 0 10 93 56 3 53 58
Hospital 6. N=70
(78 yrs)
N=93
2007,2009+1/13-4/14
86 93 68 70 64 69 0 58 72
Hospital 7. N=40
(76yrs)
N=81
2005 and 3/13-3/14/15
85 3 65 55 75 59 2 44 35
Hospital 8. N=80
(70yrs)
N=296
2003, 05, 07 and 6/13-4/14
77 91 79 20 74 70 0 42 25
Hospital 9. N=79
(75 yrs)
N=457
2001,2007+1/14-4/2015
86 89 63 20 77 70 3 75 37
Hospital 10. N=70
(74 yrs)
N=141
2007,2009+5/13-3/14
95 83 81 51 70 87 3 81 50
Hospital 11. NPR N=80
(74 yrs)
N=267
2002,05,08 + 1/13-3/14
16% and 1.02
(0%)
85 78 81 29 77 60 0 54 73
Hospital 12. N=79
(71 yrs)
N=477
2003, 05, 07+9/13-4/14
90 95 70 37 78 58 1 51 71
Hospital 13. N=79
(74 yrs)
N==266
2003, 2005,2007, Jan 13-Dec14
85 68 13 31 74 64 1 44 88
Hospital 14. N=20
N=276
7/11-8/2011
100
(Actual est. 62%)
62 0 25 78 44 0 28 14
Hospital 15. N=80
(74 yrs)
N=442
2002,03,05, Apr 13-Mar 14
15% and 1.24
4%
82 94 0 26 72 46 1 46 42
Hospital 16. N=94
(77 yrs)
N=64
12/09-4/13
19 100 NA - NA- 68 22 35 55
Hospital 17. N=70
(80)
N=196
‘07,’09, 7/13-3/14
90 88 94 27 68 58 2 54 59
Hospital 18. N=50
(77yrs
N=194
2006, 2009, 2/13-5/14
98 98 96 86 77 72 0 56 23
Hospital 19. N=100
(75 yrs)
N=187
1/10-8/12
2 24 NA 47 at 24hrs NA- 51 11 42 45
Hospital 20. N=31
(79yrs)
N=68
’06,’07, 2/13-2/14
95 77 61 50 80 56 4 32 54
Hospital 21. N=99
(80 yrs)
N=67
2/12-10/14
37 53 0 14 75 56 7 58 33
Hospital 22. N=57
(86 yrs)
N=106
1/09-6/14
0 35 11 12 77 30 14 43 17
Hospital 23. N=70
(79 yrs) N=199
2005,2011+5/13-6/14
75 69 86 56 73 22 1 54 14
Hospital 24. N=79
(74yrs)
N=315
2002, 04, 05, Jun 13-Mar 15
16% and 1.40
10%
80 89 6 29 66 38 4 45 56
Hospital 25. N=20
(74yrs)
N=178
4/12-6/2012
0 55 80 10 71 47 0 43 46
Hospital 26. N=40
(76 yrs)
N=140
2006,2010+1/13-4/14
7 65 62 65 80 60 0 58 31
Hospital 27. N=100
(80 yrs)
N=23
2004-2014
0 47 13 NA 65 39 12 39 2
Hospital 28. N=100
(77)
N=47
7/11-5/2012
0 9 0 0 80 20 3 20 33
Hospital 29. N=14
(78 yrs)
N=19
2008+3/12-3/14
7 29 7 7 46 14 3 42 0
SCAP Audit: Average process adherence by type
0
10
20
30
40
50
60
70
80
90
100
MetropolitanEnhanced
Rural Enhanced Rural Non-enhanced
Perc
en
tage
Average adherence SU/HDU/ICU
Brain Imaging within 24 hours
Neuro Obs 24 hours
Aspirin<24 hours
Physio<24 hours
Speech Path<24 hours
OT<24 hours
Use of Stroke Clinical Pathway
VTE Prophylaxis
MDT Family Meeting
Discharge on antithrombotic
Discharge on New Statin
26
SCAP audit: Process measures across 29 sites N=1788
27
The solid red line
represents access to a
SU/HDU/CCU/ICU bed
and the broken blue
line the use of a stroke
pathway
SCAP project site improvement
284 19.1 0 55 80 10 71 47 0 43
Hospital BHI
30 day
Mortality
(%)
SU/HDU
Bed
(%)
24 hr
Neuro
Ob's (%)
Clinical
P’way (%)
Swallow
test< 4
hrs (%)
%Discharged
on
A’thrombotics
Aspirin at
24 hours
(%)
Pall’
Care
(N)
% D/C
on
Statin
From a Pilot audit with poor
adherence, and a high BHI
mortality estimate, to a new
Stroke unit and now an
Acute Thrombolysis Centre.
The 2013-14 audit bridges
the inception of the new
Stroke unit but shows
substantial improvement in
process adherence
More recent audit shows
95% access to SU/HDU
and 100% antithrombotic
prescribing on discharge
• Establishing a new stroke unit.
• Patient flow review to ensure 90% of all
presenting patients are admitted to a
stroke unit
• Develop a stroke/neurology pathway
Ongoing program of ED staff education
to implement the Acute Screening of
Swallow in Stroke/TIA Training Tool
(ASSIST) for all stroke patients at
presentation.
• The development, implementation and
evaluation of a 24/7 blanket referral to
Allied Health, commencing in ED and
confirmed when the patient is admitted to
a ward bed.
• Pharmacy review of all stroke patients with
a particular emphasis on the prescribing of
anti-thrombotics and statins
• Use of local HDU beds or ambulance
bypass and hub and spoke transfer
• Specific QI for individual processes
29
Common local Quality Improvement activities
resulting from the SCVSS & SCAP
Feedback sessions engaged local clinicians and managers together, as well as
members of ASNSW and often members of the LHD executive. Local QI responses
were facilitated by a local clinician leader and Mr Mark Longworth from ACI/SCAP. Local
responses were comprehensive and new strategies shared with other sites
Results and locally agreed strategies were fed back to
LHD CE’s by the ACI Chief Executive in writing
A minority of hospitals provide organised/specialised stroke care.
At the beginning of the pilot and SCAP process there were no stroke units in
two of participating LHDs and in eastern NSW and there was no organised
stroke thrombolysis south of Campbelltown in Eastern NSW.
Since the pilot process there are four new stroke units and a new stroke service
coming on line in those areas of focus.
Three new Acute Thrombolysis Centres have come on line.
In SCAP all unenhanced sites seeing >100 strokes per year are being
enhanced or are in the process of establishing hub and spoke flows.
30
SCAP: Improving stroke unit access
Two major hospitals with higher than expected BHI 30 day mortality estimates
admitted only 50 and 60% of there ischaemic stroke patients to stroke unit beds and
access to stroke unit beds at other hospitals was lower than expected.
Audited sites with stroke units have undertaken to improve Stroke Unit/HDU/ICU
access for stroke patients to 90% or more through improved patient-flows.
A stroke pathway in a HDU has been effective at Hospital 2 where a stroke unit may
not be feasible. This approach is being considered by Hospital 21 (67 strokes pa
and 189 ks from its nearest ‘hub’), and may be widely applicable.
All participating unenhanced sites have committed to either on-site upgrades or the
use of a hub and spoke model of downstream and return flows. (Upgrades:
Hospitals 16, 19, 21, 25, and 28 and hub and spoke Hospitals 22, 27 and 29).
Ambulance bypass and facilitated transfer is being used or considered for Hospitals
21-3, 22-9, and 29-10, and Spoke A-Hospital 6, Spoke B-Hospital 1 and Spoke C-
Hospital 2, and across two participating LHDs as new units are being established.
31
SCAP: Improving stroke unit access
SCAP: Conclusions and achievements
We know that many patients do not reach a stroke unit hospital or a stroke unit bed
in a stroke unit hospital (SCAP, NSF and ACI audit data). Patients do not always
receive evidence based processes of care within stroke units (SCAP audits).
SCAP has Identified explanations for unwarranted clinical variation seen in BHI data.
ACI has taken data to the bedside. Engaging with hundreds of clinicians and
managers across NSW, providing information, expertise and support to identify and
locally address unwarranted clinical variation.
Face-to face feedback has resulted in locally developed responses to UCV.
Thirty participating sites are addressing access to desired investigations, and access
to stroke unit beds, better prescribing and the use of stroke care pathways to
improve adherence with other processes known to improve patient outcomes and
experience.
32
UCV and stroke: What is next?
Two more SCAP feedback site feedback sessions remain.
The EMR Build C including stroke bundles of care is going into testing and there is
hope that data extraction tools will be put in place to monitor site adherence with
important clinical processes.
The BHI is expected to work with ACI’s SCAP team to complete a data-linkage and
analysis between there 30 day mortality data and the audit data to validate and
improve the collection and analyses of the routinely collected data used by BHI.
The new NWAU weighting for accessing a stroke bed in NSW and the new code for
thrombolysis are expected to strongly impact on patient access.
BHI will expect to repeat the analysis of the 30 day mortality analysis which will
assess the early impact of the SCAP programme.
This ACI sponsored Stroke Forum: Reducing Unwarranted Clinical Variation for
clinician and managers involved in the stroke journey to discuss the SCAP results,
hear presentations from participating hospital sites and consider next steps in
improving stroke patient care and outcomes.33
Via plane, train, cars and ambulance (car)….
Level 4, Sage Building
67 Albert Avenue, Chatswood NSW 2067
PO Box 699
Chatswood NSW 2057
T + 61 2 9464 4666
F + 61 2 9464 4728
www.aci.health.nsw.gov.au
Daniel ComerfordDirector, Acute Care
94644602
Special Thanks
Special thanks to Mark Longworth ACI, Dominique Cadilhac and Tara Purvis (Florey
Institute), Bruce Paddock (ASNSW), Kim Sutherland and Doug Lincoln (BHI), Joseph
Descallar and Melina Gattellari (Ingham Institute), the NSW Stroke Network and the
Clinicians, Managers, Auditors and Clinical Information staff of the thirty hospitals
participating Stroke Clinical Audit Process (SCAP).
0
20
40
60
80
100
120
Hospital 2 Hospital 3 Hospital 6 Hospital 10 Hospital 17 Hospital 18 Hospital 20 Hospital 23 Hospital 26
%
Hospitals ranked from right to left by increasing estimated mortality
Process measures: A comparison of Rural Unenhanced and Metropolitan and Rural Enahcned Sites Brain Imaging*
Physio*
Speech*
OT*
Documented Swallow*
Documented Swallow**
MDT Family Meeting
Any SU/HDU
Stroke pathway
Clinical care plan
Linear (Any SU/HDU)
Linear (Stroke pathway)
SCAP: Process measures at 8 Unenhanced Rural sites N=495
37
0
20
40
60
80
100
120
Hospital 16 Hospital 19 Hospital 21 Hospital 22 Hospital 25 Hospital 27* Hospital 28* Hospital 29*
Process measures: SCAP audits of 8 unenhanced rural sites SU/HDU/ICU
24hr NeuroObs
Stroke Clinical Pathway
Swallow<24
DC Antithrombotics
Asprin<24hrs
DC on Statin
VTE Proph.
Linear (24hr NeuroObs)
Linear (Swallow<24)
Linear (Asprin<24hrs)
Linear (VTE Proph.)
Hospitals ranked from left to right by increasing estimated mortality
0
20
40
60
80
100
120
Hospital 16 Hospital 19 Hospital 21 Hospital 22 Hospital 25 Hospital 27* Hospital 28* Hospital 29*
Process measures: SCAP audits of 8 unenhanced rural sites SU/HDU/ICU
24hr NeuroObs
Stroke Clinical Pathway
Swallow<24
DC Antithrombotics
Asprin<24hrs
DC on Statin
VTE Proph.
Linear (24hr NeuroObs)
Linear (Swallow<24)
Linear (Asprin<24hrs)
Linear (VTE Proph.)
SCAP: Process measures at 9 Enhanced rural sites N=510
38
0
20
40
60
80
100
120
Hospital 2 Hospital 3 Hospital 6 Hospital10
Hospital17
Hospital18
Hospital20
Hospital23
Hospital26
%
Hospitals ranked from right to left by increasing estimated mortality
Process measures: 9 Enhanced Rural sitesBrain Imaging*
Physio*
Speech*
OT*
Documented Swallow*
Documented Swallow**
MDT Family Meeting
Any SU/HDU
Stroke pathway
Clinical care plan
Linear (Any SU/HDU)
Linear (Stroke pathway)
39
Unwarranted clinical variation and unwanted outcomes
Metropolitan hospital sites to November 2015
Ranked by increasing BHI 30 d mortality
40
Unwarranted clinical variation and unwanted outcomes
Metropolitan hospital excerpt
More GREEN squares and less RED numbers are better
