1. Is it really possible to improve quality and reduce cost in the NHS? Dr Steven Allder s.allder@nhs.net

2. My answer to the question? Yes, definitely…however…two buts A small but, and A very big BUT

3. Yes, Definitely 2:

4. Two interrelated HOW challenges… Small But Find opportunity Small But Find opportunity Very big But – mainfest opportunity

5. The small but: How to find opportunity to improve care and reduce cost Technical Analysis Principles Operational thinking – 3 rd generation systems thinking

6. YOU CAN START ANYWHERE! 1. How to find opportunity in theory 2. How to deliver results locally 3. How to deliver results more broadly Starting point is ALWAYS multiple symptoms Need a strategic framework to orientate you

7. Putting it all together: Step 1 – Where and why to start: Stroke example Strategic narrative

8. 2008 - Worst Performer in Region University Hospitals Bristol Poole Taunton And Somerset North Bristol Royal Devon And Exeter Healthcare Northern Devon Healthcare Plymouth Hospitals 118 105 Salisbury Health Care 107 Yeovil District 111 Royal United Hospital Bath 112 Royal Bournemouth and Christchurch 116 Dorset County Hospital117 South Devon Health Care Weston Area Health Royal Cornwall Hospitals Cheltenham General Hospital 102 Gloucestershire Royal Hospital 102 Great Western Hospitals Stroke in-hospital deaths by NHS hospital 4 Standardised mortality rate (percent of national average) In 2008/ 2009, PHT death rate was 18.3% higher than the national average £2,000 Loss per Patient 1.5 patients per day average £1.1 million annual loss Poor Patient & Relative Experience

9. Step 2 – Move from cost control to value creation Different model of how entities achieve success..

10. £ Challenge Value of clinical work Activity£ CC Support S S S S Current Approach Limited Scope C C Starting point Proposed Approach Clinical Value framework (1) This is very different and very ‘challenging’

11. Systems Approach: Vanguard method building on best of health care improvement science

12. Step 3 – Take a systemic view of the problem Think back to six grid model – interdependent system: critical insight.

13. In health care – there are lots of potential symptoms One department’s data from Trust Databook

14. Current approach = tackle as issues arise in isolation Proposed approach = ‘systems – value’ approach to issues

15. Key to Systems – value approach: Find value streams Then apply profound knowledge; Purpose System Variation Theory of knowledge Intrinsic Motivation

16. Step 4 – Conduct Systemic Analysis of your Sub System Appreciation of the whole..

17. Regulators System Architect (commissioners) Primary Prevention Primary Care Secondary USC Secondary SC Community Care 1 2 3 MH SC EG Te Tr R Frontline Individuals Model of Healthcare System Purpose Add value Condition Level PATIENTS Frontline Teams Frontline Management Middle Management Senior Management Direct value creation Indirect value creation

18. Step 5: Find strategic ‘entry’ patient demand stream Population segmentation exercise Programme Budgeting

19. Primary Prevention Primary Care Secondary USC Secondary SC Community Care 1 2 3 MH SC EG Te Tr R Model of Healthcare System: Demand streams

20. 1-Apr-061-May-061-Jun-061-Jul-061-Aug-061-Sep-061-Oct-061-Nov-06 1-Dec-061-Jan-07 1-Feb-071-Mar-071-Apr-07 1-May-07 1-Jun-07 1-Jul-07 1-Aug-071-Sep-071-Oct-071-Nov-07 1-Dec-07 1-Jan-08 1-Feb-08 Mean Number of admissions Mean has been 1.5 admissions per day over the past two years UCL* Inpatient demand stream Stroke Patients Admitted per Day

21. Step 6: Find key consumption stream Understand authentic cause of variation in performance: 1. Demand sub-type 2. Stage of care 3. Step of care 2121

22. High ( 8+) Med (3-7) Low (<3) 2009 499 (84) 83 (14) 18 (3) ALOS Days Number of bed days (percent) MeanLOS ALOS Days Consecutive patients Length of stay for patients with ALOS ≥8 days 22 Quality: Interaction of streams Identify Hot Spots of resource consumption (Ideally, quality then cost)

23. Primary Prevention Primary Care Secondary USC Secondary SC Community Care 1 2 3 MH SC EG Te Tr R Model of Healthcare System: Consumption streams

24. Step 7: Dissect strategic consumption stream Understand authentic cause of variation in performance: 1. Homogeneous demand sub-type 2. Stage of care driving variation 3. Step of care driving variation 2424

25. Quality Grid guided case note review: Emergent element of process Presenting complaint/D iagnostic sub -division

26. * Rehabilitation Stroke Unit ** Frail patients were defined as having medical complexity index of 3, 4, or 5 on a 1-5 scale. 0=No systemic disease other than primary diagnosis, 1=Premorbid, inactive, and or irrelevant systemic disease, 2=Active, relevant systemic disease not limiting function, 3=Active, systemic disease limiting function, 4=Active, systemic disease severely limiting function, 5=Moribund / terminal intermediate Home (23%) RSU (13%) RSU (11%) RSU or convalescence (20%) RSU* or convalescence (17%) Well (47%) Frail** (53%) MildModerateSevere Clinical stroke size Patient status pre- stroke Preferred place of discharge for 6 subgroups of patients 3 (percent of total) Pathway redesign required (16%) 7a. Which Demand sub type (s): Six types of patients were defined based on patient status pre-stroke and the size of the stroke

27. Pathway redesign required (16%) Home (23%) RSU (13%) RSU (11%) RSU or conval- escence (20%) RSU or conval- escence (17%) Well (47%) Frail (53%) MildModerateSevere Clinical stroke size Patient status pre- stroke Step 7b: Which Stage of care Target Group 1 – Redesign Pathway: Discharge Target Group 2 – Improve Operational Rigour: Rehab

28. Pathway redesign required (16%) Home (23%) RSU (13%) RSU (11%) RSU or conval- escence (20%) RSU or conval- escence (17%) Well (47%) Frail (53%) MildModerateSevere Clinical stroke size Patient status pre- stroke Step 7c: Which Step of care Target Group 1 – Redesign Pathway: Whole process Target Group 2 – Improve Operational Rigour: ASU-RSU

29. Step 8: Redesign steps The pathway was then redesigned for the key segment (frail patients with severe stroke), and operational improvements were initiated in the RSU for four other patient segments

30. Structuring thinking about designing the specific step care: frail 1.Frail Elderly RIP 2.Morning handover 4.Expectant pathway 3.Very intimate 5.Protocol 7.Team 6.LCP Hospice 6.LCP Hospice 8. Costs far less

31. Step 9: Implement Redesign The pathway was then redesigned for the key segment (frail patients with severe stroke), and operational improvements were initiated in the RSU for four other patient segments

32. Severe Stroke in Frail Patients  Highest resource consumption  75% of beds were used by the frail patients pre-stroke  Highest variability in bed occupancy & long length of stay  Driven by a lack of systematic care planning  Care not well-matched to patients  Variable treatment and feeding processes, not aligned with patient and relative preferences

33. RSU Operational Rigour (1)  No frail patients with severe strokes are sent to RSU  Active decision for frail patients with moderate stroke  Based on clear triage rules and input from acute care providers, relatives and patients  Previously well patients with moderate or severe strokes go to the RSU  Rigorous monitoring is used to determine when patients can be sent home with enhanced community resources (early supportive discharge) or to long-term placement (e.g., nursing home)

34. Step 10: Measure Impact The pathway was then redesigned for the key segment (frail patients with severe stroke), and operational improvements were initiated in the RSU for four other patient segments

35. In under a year, access to and use of the Acute Stroke Unit has become more efficient 3 -12% p.a. FebJanDecNovOctSepAugJulJunMayApr -6% p.a. FebJanDecNovOctSepAugJulJunMayApr +7% p.a. JunMayAprMarFebJanDecNovOctSepAugJulJunMayAprMar 20092010 Percentage of patients spending at least 90% of their time in the ASU *,3 Percentage 20092010 Time required for transfer to ASU 3 Hours 20092010 Average LOS in ASU 3 Days * This is one of the major indicators in the UK National Stroke Audit; if patients are not spending time in the stroke unit, they are either in the A&E or the medical assessment unit, likely not getting the most appropriate care

36. June 2009April 2009 Acute beds at Derriford Hospital Rehab beds at Mount Gould May 2009April 2009 Net acute benefits Reimbursement level: £4k per patient New cost of care: £3k per patient, Savings: £1k 17 beds released, implying net savings of 11% across system Lots of Beds Saved…Permanently

37. Step 11:Repeat…

38. An offer of hope