Determining the Link between Structure Process and Outcome

Determining the Link between Structure, Process, and Outcome in Stroke Is It Realistic? What are the Right Metrics? Eric D. Peterson, MD, MPH Professor of Medicine Vice Chair of Quality, Dept of Medicine Director of CV Research Duke Clinical Research Institute (DCRI) Author Disclosure: Research support from BMS-Sanofi, Merck, Eli Lilly, Ortho Mc. Neil

What is Quality? “Degree to which health care services increase the likelihood of desired health outcomes and are consistent with current professional knowledge” l Are we doing the right things? l Are our patients better off for it?

Domains of Quality Structure-Process-Outcome Structure Process Outcome Each of the components in the model has a direct influence on the next component in the model Donabedian A, Milbank Quarterly, 1966

Structure Role of Specialty Stroke Care n Caplan (2003) – Stroke care should be managed by neurologist n Lee (2003) – Stroke patients should be cared for by stroke subspeciality n Kazmierski et al (2004) – Stroke care should be holistic and include (Neurologist) Caplan L. Stroke is best managed by neurologists. Stroke. 2003; 34: 2763. Lee KR. Stroke is best managed by neurologists: Battle of the titans. Stroke. 2003; 34: 2764 -2765. Kazmierski R, Pawlak MA, Kozubski W. Who should care for stroke patients? Stroke. 2004; 35: e 85 -e 86.

Its not Individuals but Teams that Matter!!

Role for Team-based Stroke Care? There is evidence that specialized stroke teams improve outcomes Bershad EM, Feen ES, Hernandez OH, Suri MF, Suarez JI. Impact of a specialized neurointensive care team on outcomes of critically ill acute ischemic stroke patients. Neurocrit Care. 2008; 9(3): 287 -292. Adams H, Adams R, Del Zoppo G, Goldstein LB. Guidelines for the early management of patients with ischemic stroke: 2005 guidelines update a scientific statement from the Stroke Council of the American Heart Association/American Stroke Association. Stroke. Apr 2005; 36(4): 916 -923. Deborah J. Webb DJ Fayad PF, Wilbur C, Thomas A, Brass LM. Effects of a Specialized Team on Stroke Care: The First Two Years of the Yale Stroke Program. Stroke. 1995; 26: 1353 -1357.

Need for Systems of Care for All Not everyone has the luxury of a nearby primary stroke center Spoke Hub Neurologist Spoke

Stroke Systems of Care Regionalization

Telestroke: Bringing the Specialist to the Patient Video may and improves upon consultation by allowing… n Visualization and interaction with patient for more accurate assessment n Ability to obtain a valid NIHSS score n Ability to view a CT scan in real time for reliable diagnosis n Sharing patient records with remote physician Vaishnav. Expert Rev Neurother. 2007; 7: 913 -914. Shafqat. Stroke. 1999; 30: 2141 -2145

Stroke Systems Controlled Trial TEMPIS Study – Lancet Neurology 2006 • 10 community hospitals in Bavaria – 5 intervention, 5 control – Matched in size, infrastructure • Intervention – Stroke wards and medical staff training – 24/7 telemedicine consultation from academic centers (both ED and inpatient) • July 2003 – March 2005, 3122 ischemic and hemorrhagic stroke patients • Patients dead or disabled 3 months after stroke 43. 6% vs 54. 0%, p<0. 001, OR for poor outcome 0. 61, 95% CI. 51 -. 73

Do Stroke Centers Reduce Mortality for Patients with Acute Ischemic Stroke? An Instrumental Variable Analysis n Ying Xian et al AHA QCOR YI competition 2010 n Studied 32, 783 NYS stroke patients l 50% treated at primary stroke centers l Used distance to stroke center as IV n Results: l Adjusted mortality lower at primary stroke centers Xian Y et al Abstract at AHA QCOR 2010

Why Structural Measures are Imperfect Associations are just that…. Ross et al. New Engl J Med, 2010.

Stroke Process of Care: Performance Measures • % of acute ischemic stroke (IS) pts who arrive at ED within (2 hrs) of onset of symptoms who receive IV t-PA within 3 hrs • % IS or TIA pts who receive antithrombotic medication within 48 hours of hospitalization • % of IS or TIA pts discharged on antithrombotics • % of IS or TIA patients with atrial fibrillation who are discharged on anticoagulation therapy unless a contraindication • % of pts at risk for DVT who received DVT prophylaxis by Day 2 • % of IS or TIA patients with LDL>100 mg/d. L or on cholesterol reducer prior to admission who are discharged on cholesterolreducing drugs • % of smokers who receive smoking cessation advice or meds at discharge GWTG- Fact Sheet (2006)

How are We Doing on Process 1, 000 Patients Strong!

Improving Care Process In GWTG-Stroke

Opportunity to Improve Timeliness of IV rt-PA Door-to-IV rt-PA within 60 minutes GWTG-Stroke Database, data on file DCRI

Unequal Access to Evidence-based Stroke Practices Gender Reeves M et al Stroke. 2009; 40: 1127 -1133. Race / Ethnicity Schwamm L et al Circulation. 2010; 121: 1492 -1501. )

Performance Measurement: Its Not as Easy as You Think n Do we have right metrics? l Measuring “what’s easy” vs “what’s important” n What is precision of our performance tools? l Sample size issues l Need to risk adjust performance metrics? l Composite measure complexity n Can we change it? l P 4 P vs provider led efforts l Where do we go from here?

Evolving View of Quality Care: Importance of Longitudinal Measures Post AMI Trajectory 1 (T 1) Relatively healthy adult Assessment of Preferences Focus on: • Quality of Life • Functional Status • 20 Prevention Strategies • Rehabilitation • Advanced care planning Population at Risk 10 Prevention (no known CAD) 20 Prevention (CAD no prior AMI) Acute Phase Post Acute/ Rehabilitation Phase PHASE 2 PHASE 3 20 Prevention (CAD with prior AMI) Advanced Care Planning PHASE 4 PHASE 1 Staying Healthy Getting Better Episode begins – onset of symptoms Living w/ Illness/Disability (T 1) Coping w/ End of Life (T 2) Post AMI Trajectory 2 (T 2) Adult with multiple co-morbidities Focus on: • Quality of Life • Functional Status • 20 Prevention Strategies • Advanced Care Planning • Advanced Directives • Palliative Care/Symptom Control Episode ends – 1 year post AMI

Stroke Outcomes Mortality

Potential to Improve? Variation in Hospital Stroke Mortality Outcomes, Adjusted Event Rates Distribution of Hospital Event Rates Mean ± STD 10 th 25 th Median 75 th 90 th In-Hospital 5. 7 ± 2. 2% 2. 8% 4. 6% 6. 0% 6. 9% 8. 2% 30 -Day 13. 9 ± 3. 5% 9. 8% 12. 1% 14. 2% 15. 6% 17. 8% 90 -Day 20. 1 ± 4. 2% 15. 3% 18. 2% 20. 2% 22. 2% 25. 1% 1 -Year 30. 9 ± 4. 6% 25. 2% 28. 6% 31. 0% 33. 0% 36. 3%

Stroke Outcome Issues n Low numbers of events n Need for risk adjustment l NIH stroke scale most predictive yet incomplete n What is a good stroke outcome? l Is discharge to SNF with severe disability worse? n What about other outcomes l Readmission l Recurrent events l Functional Recovery? Weisscher N, Vermeulen M, Roos YB, de Haan RJ. What should be defined as good outcome in stroke trials; a modified Rankin score of 0 -1 or 0 -2? J Neurol. 2008 Jun; 255(6): 867 -74

The Stroke Quality Agenda… Where Do We Go From Here… Structure Implementation Process Outcomes

New Science: New Therapeutics High Throughput Screening: (HTS) n l n allows rapid screening of a high #’s of chemicals to find an active compound. The starting points for understanding the role of a particular biochemical process Genome-wide Association (GWAS) l. In-depth characterization of patients genes to identify those factors that are different in those with trait of interest Study Perturbation of Biological Systems n l. Causing an experiments disruption of a system to better understand its properties

New Science…Implementation Science n High Throughput Screening: (HTS) (AKA National Clinical Registries) l l n allows rapid screening of hundreds of centers to find those with outstanding “Positive or Negative Deviance” GWAS (AKA Qual/Quant Research) l In-depth characterization of hospital structure and process to identify those factors associated with better outcomes n. Controlled l. Natural Perturbation Experiments l. Cluster randomization

Conclusions n Structure l Current: Specialty MD, team, hospital, system l Future: Understand HOW? n Process l Current: established stroke performance metrics l Future: study closely the link to outcomes; Need broader metrics: n Outcomes l Current: In-patient mortality l Future: Patient centric, longitudinal measures