Acute myocardial infarction: Tracking patients journeys and outcomes in a complex, acute healthcare system

Transcription

1 Acute myocardial infarction: Tracking patients journeys and outcomes in a complex, acute healthcare system NHS Greater Glasgow and Clyde, Golden Jubilee National Hospital, University of Glasgow, DataLab and AstraZeneca UK This analysis has been performed by AstraZeneca as part of the Joint Working project between NHS Greater Glasgow and Clyde (GGC), Golden Jubilee National Hospital (GJNH), GGC Managed Clinical Network and AstraZeneca UK Ltd.

2 Background Reducing cardiovascular mortality and improving outcomes is a key focus for NHS Scotland Scottish hospitals do not submit data to Myocardial Ischaemia National Audit Programme (MINAP) We lack of contemporary data on service delivery and outcome in our patients.

3 Background In Glasgow, electronic patient records have been implemented in place of paper records National Advisory Group for Heart Disease pilot e-registry NHS Greater Glasgow and Clyde (GGC), Golden Jubilee (GJNH), Glasgow Managed Clinical Network (GMCN) and AstraZeneca UK (AZ) = Joint Working Agreement MCN involves 6 acute hospitals in NHS GGC (population 1.2 million) and the regional hub for invasive cardiology (GJNH)

4 Pathway Map 1 Local A&E 2 Local Hospital 5 3 GJNH 6 9 Home 4 Pathway Beginnings 1 Emergency Direct to GJNH 2 Local A&E to GJNH 3 Direct Transfer from local Hospital to GJNH 4 Elective GJNH admission following discharge and referral from local hospital 5 Local Hospital only (no referral or GJNH admission) 6 Elective GJNH admission only (no referral from local hospital or local admission) Pathway Endings A Discharged home (from local or GJNH) B Transferred to local hospital before discharge C Death in hospital (local or GJNH)

5 Pilot Objectives 1. Design and create a linked CV patient registry for patients treated across NHS GGC and GJNH using routinely collected electronic healthcare records 2. Complete a service evaluation focusing upon the delivery 12 months of ACS services managed across NHS GGC and the GJNH 3. Establish a robust legacy of quality data collection for CV patients that can be used for research and service improvement purposes into the future 4. Project steering committee to make recommendations for optimal service design and local delivery plan to GMCN and National Advisory Committee for Coronary Heart Disease

6 Key Early Activities Project steering group - NHS GGC, GJNH, GMCN & AstraZeneca UK Permissions from all relevant regional and national bodies Joint Working Framework little financial resource needed Records extracted from three sources of electronic patient records from NHS GGC and GJNH; initially for 3 month period to enable testing of extract specification and process Pilot registry created by Safe Haven at University of Glasgow to link records extracted Methodology for analyses and reports designed by AZ with support of NHS clinical and technical teams

7 Pathway Definitions Pathway Description Admission and Discharge Criteria Additional Criteria 1 Direct (emergency) to GJNH GJNH Admission method either 'Direct to Cardiac Centre or 'Air ambulance transfer' No referral prior to GJNH admission. Must have valid discharge recorded 2 Via local A&E to GJNH 3 4 Local hospital inpatient admission with direct transfer to GJNH Local hospital inpatient admission, discharged home with referral to GJNH 5 Local hospital treatment only 6 Elective GJNH treatment only GJNH Admission method either 'Ambulance to local A&E or 'Self-referral to local A&E' GJNH Admission method either 'Transfer from ward in local hospital or 'Already in cardiac centre' GJNH Admission method either 'Elective admission or 'Direct to cardiac centre' Admission direct to local hospital, discharged home or within local hospital GJNH Admission method either 'Elective admission or 'Direct to cardiac centre. Discharge method either 'Discharged home or 'Inpatient death' No referral prior to GJNH admission. Must have valid discharge recorded Local Hospital admission and referral prior to GJNH admission. Must have valid discharge recorded. Local Hospital admission and referral prior to GJNH admission. Local hospital discharge must be Home. No referral for further activity after local hospital admission or evidence of GJNH admission. Must have valid discharge recorded. No referral prior to GJNH admission or evidence of local hospital admission. Must have valid discharge recorded.

8 Summary of 12 month Data TRAKCARE: 5445 records SCI Gateway: 6651 records CATHI: 7157 records Mortality: 862 records Excluded Spells with incomplete data* Non-IHD Final Diagnosis Ischaemic Heart Disease (IHD) Linked Dataset Unique Patients: 3186 Distinct Complete Spells: 3438 *Incomplete Data: Pathway cannot be assigned due to missing information, normally due to start or end of pathway falling outside of current 12 month data window or patient attending from Health Board other than NHS GGC, therefore no TRAKCARE data available in current extract. Also includes patients where SCI referral was made but there is no evidence of any clinical activity in dataset. Analysis performed by AstraZeneca as part of the Joint Working project between NHS Greater Glasgow and Clyde (GGC), Golden Jubilee National Hospital (GJNH), GGC Managed Clinical Network and AstraZeneca UK Ltd.

9 Example Outputs 9 Asset ID: 594, Date of Prep: November 2014

10 NSTEMI Summary MINAP National Audit Annual Report NHS GGC Health Board Summary Oct 2013-Sept (63%) referred for or had angiography 451 (65%) admitted via emergency services or selfpresentation 200 (28%) admitted via transfer for specific treatment 46 (7%) admitted via another or unknown method 48 (11%) angiography within 24hr of admission 38 (8%) angiography 24-48hr after admission 64 (14%) angiography 48-72hr after admission 63 (14%) angiography 72-96hr after admission 238 (53%) angiography >96 hr after admission Analysis performed by AstraZeneca as part of the Joint Working project between NHS Greater Glasgow and Clyde (GGC), Golden Jubilee National Hospital (GJNH), GGC Managed Clinical Network and AstraZeneca UK Ltd.

11 Non ST-Elevation Myocardial Infarction (NSTEMI)

12 NSTEMI Patients: Spells by Care Pathway 1 n=30 (3%) Local A&E 2 n=6 (1%) Local Hospital 5 n=403 (37%) Pathway Definition 1 Emergency Direct to GJNH 2 Local A&E to GJNH 3 Local hospital inpatient admission with direct transfer to GJNH 4 Local hospital inpatient admission, discharged home with referral to GJNH 5 Local Hospital only 6 Elective GJNH only 3 n=415 (38%) Home 4 n=200 (18%) GJNH 6 n=46 (4%) N=1100 Analysis performed by AstraZeneca as part of the Joint Working project between NHS Greater Glasgow and Clyde (GGC), Golden Jubilee National Hospital (GJNH), GGC Managed Clinical Network and AstraZeneca UK Ltd.

13 Mean Days in Hospital 20.0 NSTEMI Patients: Mean Days in Hospital by Pathway N= (median=1) 1 Direct to GJNH (n=30) 8.0 (median=4) 2 Local A&E to GJNH (n=6) 7.3 (median=6) 3 Direct Transfer (n=415) 3.4 (median=3) 4 Discharged Home with Referral (n=200) 11.4 (median=6) 5 Local Hospital Only (n=403) 0.8 (median=0.5) 6 Elective GJNH Only (n=46) Care Pathway Notes: If a patient has an admission that lasts less than one day (i.e. Admitted and discharged on the same date) then for the purposes of calculation of days in hospital they have been assigned a duration of 0.5 days for that admission. Analysis performed by AstraZeneca as part of the Joint Working project between NHS Greater Glasgow and Clyde (GGC), Golden Jubilee National Hospital (GJNH), GGC Managed Clinical Network and AstraZeneca UK Ltd.

14 Percent of Spells in the Pathway NSTEMI Patients: Percentage of Spells Receiving Angiography and Percentage Receiving PCI By Pathway 100.0% 99.1% 100.0% 100.0% 100.0% 100.0% 90.0% 83.3% 87.0% 80.0% 70.0% 60.0% 50.0% 49.8% 60.0% 54.5% 56.5% N= % 30.0% 20.0% 10.0% 36.0% Angiography PCI 0.0% Any Pathway (n=697) 1 Direct to GJNH (n=30) 2 Local A&E to GJNH (n=6) Care Pathway 3 Direct Transfer (n=415) 4 Discharged Home with Referral (n=200) 6 Elective GJNH Only (n=46) Analysis performed by AstraZeneca as part of the Joint Working project between NHS Greater Glasgow and Clyde (GGC), Golden Jubilee National Hospital (GJNH), GGC Managed Clinical Network and AstraZeneca UK Ltd.

15 Percent of Patients NSTEMI: Distribution of Days in Hospital According to the Pathway N= < >21 1.Direct to GJNH Local A&E to GJNH Direct Transfer Discharged Home with Referral Local Hospital Only Elective GJNH Only Days in Hospital Analysis performed by AstraZeneca as part of the Joint Working project between NHS Greater Glasgow and Clyde (GGC), Golden Jubilee National Hospital (GJNH), GGC Managed Clinical Network and AstraZeneca UK Ltd.

16 Percent of Spells in the Pathway NSTEMI Patients: Percentage of Spells Receiving Angiography and Percentage Receiving PCI By Pathway 100.0% 99.1% 100.0% 100.0% 100.0% 100.0% 90.0% 83.3% 87.0% 80.0% 70.0% 60.0% 50.0% 49.8% 60.0% 54.5% 56.5% N= % 30.0% 20.0% 10.0% 36.0% Angiography PCI 0.0% Any Pathway (n=697) 1 Direct to GJNH (n=30) 2 Local A&E to GJNH (n=6) Care Pathway 3 Direct Transfer (n=415) 4 Discharged Home with Referral (n=200) 6 Elective GJNH Only (n=46) Analysis performed by AstraZeneca as part of the Joint Working project between NHS Greater Glasgow and Clyde (GGC), Golden Jubilee National Hospital (GJNH), GGC Managed Clinical Network and AstraZeneca UK Ltd.

17 Mean Days in to Angiography NSTEMI Patients: Mean Days from Acute Admission to Angiography (Pathways 1-4) N=651 NHS Standard of Care < 72 hours (median=6) Any Pathway (n=651) 0.6 (median=0) 1 Direct to GJNH (n=30) 0.3 (median=0) 2 Local A&E to GJNH (n=6) 5.0 (median=4) 3 Direct Transfer (n=415) 26.0 (median=23) 4 Discharged Home with Referral (n=200) Care Pathway Analysis performed by AstraZeneca as part of the Joint Working project between NHS Greater Glasgow and Clyde (GGC), Golden Jubilee National Hospital (GJNH), GGC Managed Clinical Network and AstraZeneca UK Ltd.

18 Percent of Patients Referred to GJNH NSTEMI Patients: GRACE Score for Patients Referred to GJNH from Local Hospital (Pathways 3 and 4) 60% 50% GRACE >140 criterion for referral N=615 40% 30% 26% 3% 4: Discharged Home with Referral 20% 38% 3: Direct Transfer 10% 27% 0% <140 (n=330) 140 (n=252) 3% 2% No Score (n=33) GRACE ACS Risk Score Recorded at Time of Referral Analysis performed by AstraZeneca as part of the Joint Working project between NHS Greater Glasgow and Clyde (GGC), Golden Jubilee National Hospital (GJNH), GGC Managed Clinical Network and AstraZeneca UK Ltd.

19 Mortality Rate (Patients with Data) 20.0% 18.0% 16.0% 14.0% 12.0% 10.0% 8.0% 6.0% 4.0% 2.0% 0.0% NSTEMI Patients: All Cause Mortality (Unadjusted) N= % 3.4% 5.6% 7.9% 14.2% In Hospital (n=1100) 30 Days (n=941) 90 Days (n=941) 180 Days (n=771) Days From Most Recent Admission 365 Days (n=288) NB Only for patients with sufficient follow-up for each timeframe Analysis performed by AstraZeneca as part of the Joint Working project between NHS Greater Glasgow and Clyde (GGC), Golden Jubilee National Hospital (GJNH), GGC Managed Clinical Network and AstraZeneca UK Ltd.

20 Mortality Rate (Patients with Data) NSTEMI Patients: Mortality of Cardiovascular Cause (Unadjusted) 14.0% 12.0% 10.0% 8.0% 6.0% 4.0% N= % 0.0% 2.3% 2.6% 4.0% 4.7% 8.3% In Hospital (n=1100) 30 Days (n=941) NB Only for patients with sufficient follow-up for each timeframe 90 Days (n=941) 180 Days (n=771) Days From Most Recent Admission 365 Days (n=288) Analysis performed by AstraZeneca as part of the Joint Working project between NHS Greater Glasgow and Clyde (GGC), Golden Jubilee National Hospital (GJNH), GGC Managed Clinical Network and AstraZeneca UK Ltd.

21 % of Deaths in Patients with Data NSTEMI Patients: Mortality Due to Ischaemic Heart Disease (Unadjusted) 12.0% 10.0% N= % 6.0% 4.0% 2.0% 0.0% 1.9% 2.3% 3.3% 3.9% 6.9% In Hospital (n=1100) 30 Days (n=941) NB Only for patients with sufficient follow-up for each timeframe 90 Days (n=941) 180 Days (n=771) Days From Most Recent Admission 365 Days (n=288) Analysis performed by AstraZeneca as part of the Joint Working project between NHS Greater Glasgow and Clyde (GGC), Golden Jubilee National Hospital (GJNH), GGC Managed Clinical Network and AstraZeneca UK Ltd.

22 Mean Days in Hospital NSTEMI Patients: Mean Days in Hospital by Originating Hospital (median=5) Glasgow Royal Infirmary (n=289) 8 (median=6) Inverclyde Royal Hospital (n=93) 7 (median=4) Royal Alexandra Hospital (n=195) 9 (median=5) Southern General Hospital (n=59) 10 (median=5) Victoria Infirmary (n=161) 8 (median=4) Western/Gartnavel Hospitals (n=193) NHS GGC Hospital of Origin Vale of Leven Hospital not shown due to low patient numbers (n=2 ) Analysis performed by AstraZeneca as part of the Joint Working project between NHS Greater Glasgow and Clyde (GGC), Golden Jubilee National Hospital (GJNH), GGC Managed Clinical Network and AstraZeneca UK Ltd.

23 Referring Hospital Referring Hospital NSTEMI Patients Referred with Direct Transfer: Mean Time from Admission to Referral and from Referral to Treatment by Referring Hospital NSTEMI Patients Referred via Discharge Home: Mean Time from Admission to Referral and from Referral to Treatment by Referring Hospital Time to Referral Western/Gartnavel Hospitals (n=78) 3 3 Time to Referral Western/Gartnavel Hospitals (n=33) 2 18 Time from Referral to Treatment Victoria Infirmary (n=39) 3 4 Time from Referral to Treatment Victoria Infirmary (n=44) 3 22 Southern General Hospital (n=22) 2 3 Southern General Hospital (n=14) 2 20 Royal Alexandra Hospital (n=81) 2 4 Royal Alexandra Hospital (n=47) 2 30 Inverclyde Royal Hospital (n=45) 3 3 Inverclyde Royal Hospital (n=5) 3 14 Glasgow Royal Infirmary (n=141) 2 3 Glasgow Royal Infirmary (n=50) Mean Days Mean Days Analysis performed by AstraZeneca as part of the Joint Working project between NHS Greater Glasgow and Clyde (GGC), Golden Jubilee National Hospital (GJNH), GGC Managed Clinical Network and AstraZeneca UK Ltd.

24 Moving forward Aims and objectives for next 12 months 1. Regional ACS eregisty with automated data collation and reporting systems 2. Development and exploration of expanded dataset to identify new group or pathways of care of clinical significance, e.g. Ambulance Service 3. Investigate additional data sources that could be integrated into the registry to improve the richness of the data

25 Innovation in Data Science Key tenet of SG policy as described in CMO Realistic Medicine publication is Reducing Variation but it first needs to be measured Provide a paradigm shift by developing a timely, on demand real time reporting tool Immediate focus is GGC & GNJ HB but adaptable nationally Develop reporting to suit Scottish perspective / needs Use different data science techniques to explore potential groups or patterns of care of clinical interest Link to wider datasets to provide richer more detailed information on care and long term outcomes

26 Data reporting Real-time on demand reporting NHS datasets TRAKCare SLA for data supply NHS Safe Haven NHS Scottish Ambulance Service SCI-Gateway CATHi Lab Mortality eregistry Dataset ETL & process of data Server 1- Linked identified datasets Approved NHS Staff using existing software and access rights

27 LPAC Approved Data extracts Data Science / Link to wider datasets NHS datasets TRAKCare Scottish Ambulance Service SCI-Gateway SLA for data supply eregistry Dataset NHS Safe Haven ETL & process of data Server 1- Linked identified datasets CATHi Lab Mortality Other datasets Server 2- Linked anonymised datasets Approved researchers in controlled Safe Haven environment

28 Innovation in Data Science Apply novel techniques (e.g. Bayesian techniques, machine learning, pattern recognition etc) to derive new insight in support of clinical decisionmaking

29 Long term outcomes University of Glasgow MVLS Industrial Partnership PhD with AstraZeneca Aim: to explore the association of socio-economic status with the process and long-term outcomes of treatment for suspected acute MI Use data linkage to undertake epidemiological research to better understand the patient journey and factors which influence treatment decisions and outcomes.

30 The Learning Health System

31 Next steps 1. Share the outputs NHS Scotland 2. National Steering Group in Scottish Cardiac Society 3. Support other Boards to develop their own e-registry, potentially through Safe Haven 4. National e-registry 5. Beyond MINAP all cardiovascular hospitalisation.

32 Learnings 1. Both clinical and technical representation in steering group is essential to ensure understanding and delivery of requirements 2. Strong project management needed to drive project forwards to meet milestones 3. Validation of data (both underlying source and analysis assumptions) is worth investing in 4. It is possible to create a quality patient registry from existing data sources, without the need for additional data collection or excessive use of NHS resources 5. Joint Working between the NHS, academia, and the pharmaceutical industry can facilitate the delivery of projects that are valuable to the NHS and patients

33 Thank you for your attention