1. Critical Decisions in the Emergency Department University of Pennsylvania: Brendan G. Carr, MD MS Sage Myers, MD MS Scott Lorch, MD MS Patrick Reilly, MD Dylan Small, PhD Charles C. Branas, PhD Agency for Healthcare Research & Quality Ryan Mutter, PhD

2. How do we design and measure the emergency care system? (Trauma as a case study) University of Pennsylvania: Brendan G. Carr, MD MS Sage Myers, MD MS Scott Lorch, MD MS Patrick Reilly, MD Dylan Small, PhD Charles C. Branas, PhD Agency for Healthcare Research & Quality Ryan Mutter, PhD

3. Disclosures Federal research funding  AHRQ, NICHD, CDC, NINDS  www.traumamaps.org www.traumamaps.org  www.strokemaps.org www.strokemaps.org AHA research funding  NRCPR/GWTG National Quality Forum  RECS Steering Committee HHS/ASPR Senior Policy Advisor  I am not appearing in this role today

4. Conceptual Framework Ambulatory Care Sensitive Conditions  Conditions for which good outpatient care can potentially prevent the need for hospitalization, or for which early intervention can prevent complications or more severe disease. Emergency Care Sensitive Conditions  Conditions for which rapid diagnosis and early intervention in acute illness or acutely decompensated chronic illness improves patient outcomes

6. Background: The Volume-Outcome relationship 12 surgical procedures12 surgical procedures  CABG, AAA, TURP, etc. 1500 hospitals1500 hospitals Procedures Mortality =

8. Standardized Mortality 70% 65% 60% 55% 50% 20406080100120 45% Cardiac Arrest Patients Admitted to ICU/year Hospitals that treated over 50 pts/year had lower mortality Cardiac Arrest Mortality

10. Background: The time-outcome relationship

12. Sepsis

15. STEMI

16. 30+% of STEMI patients get no reperfusion therapy

18. 3% of ischemic strokes treated at TJC certified centers 3-8.5% receive rt-PA

19. Sepsis

20. 5-7% of EDs perform EGDT

22. 26% of physicians have used hypothermia (ever)

23. Volume Time

25. the organization of a system for the delivery of health care within a region to avoid costly duplication of services and to ensure availability of essential services. –Mosby’s medical dictionary What is regionalization?

26. Regionalized Trauma Care

27. Prehospital triage

28. The Trauma Model – Inventory

29. Access to trauma care

30. Trauma care outcomes

31. Trauma Model. All success? 27,130,283 injuries treated in US hospitals in 2006  32% in trauma centers  68% in non-trauma centers Severely injured patients (ISS>15)  - More likely to be treated in trauma centers  (51.3% TC vs. 48.7% nTC, p<0.001) Critically injured patients (ISS>25)  - More likely to be treated in non-trauma centers  (41.6% TC vs. 58.4% nTC, p<0.001)

32. Research questions with policy implications Have we improved population outcomes for injury?  1. In a nationally representative analysis – Do trauma centers save lives?  2. What is the relationship between access to trauma care and injury outcomes? (supply and demand)

33. Research questions with policy implications (What can understanding population outcomes for trauma teach us about examining other systems created to focus on unplanned critical illness?)  Stroke  STEMI  Cardiac arrest…

34. Q1. In a nationally representative analysis – Do trauma centers save lives? Population:  All injured patients treated at trauma centers and non- trauma centers in the US Data:  Nationwide Emergency Department Sample (HCUP)  Trauma Center Level (American Trauma Society) Geography  Patient location, hospital location (US census, ArcGIS) Prehospital transport time estimates  - empirically derived & arcGIS network analyst

35. Q1. In a nationally representative analysis – Do trauma centers save lives? Analysis  Logistic regression  Survey weights  Confounders –Age, injury severity, comorbid conditions, region, insurance, hospital size, teaching status, hospital ownership, (prehospital time)  Sub groups –Severely injured, penetrating, blunt, age > 55, only patients surviving to admission

36. Characteristics of hospitals with ED encounters for injury - 2009 CharacteristicPercent Level 1 Trauma Center12.96 Level 2 Trauma Center12.29 Level 3 Trauma Center9.63 Non-trauma Center65.12 Public Hospital15.73 For-profit Hospital14.24 Not-for-profit Hospital70.03 Teaching Hospital35.27 Large Hospital53.66 Medium Hospital28.82 Small Hospital17.52 Urban Hospital72.96

37. Characteristics of injured patients - 2009 VariableLevel 1Level 2Level 3 Non- trauma Demographics Age (average)41.3644.4043.7544.11 Male (percent)57.6452.6451.3850.66 Medicare (percent)14.7720.3420.5720.33 Medicaid (percent)14.7712.4111.9713.26 Private insurance (percent)34.7637.5738.2337.11 Uninsured (percent)25.7019.6719.1119.73 Other payer (percent)10.0010.0110.129.57 Comorbidities Has no comorbidities (average)71.2469.3771.0175.24 Has comorbidities (average)28.7630.6328.9924.76 Injury Characteristics Injury Severity Score (average)3.623.202.712.54 Severe injury, ISS > 15 (percent)3.692.291.020.64 Blunt trauma (percent)54.9657.7057.7852.98 Penetrating trauma (percent)11.5611.3211.3311.11 Self-inflicted (percent)1.160.880.690.57 Assault (percent)9.455.514.413.98

38. Relation between treatment at a level 1 or 2 trauma center and death PopulationUnadjusted Beta (t-statistic) AdjustedAdjusted with Instrument All injuries.00512 ** (12.19).00109 ** (4.81) -.00194 ** (-3.07) Injuries with ISS > 15.05859 ** (17.05).00964 ** (2.88) -.04705 * (-2.54) Blunt trauma.00583 ** (12.85).00135 ** (4.66) -.00294 ** (-3.08) Penetrating trauma.00913 ** (8.30).00225 ** (4.16) -.00009 (-0.07) Aged > 55.01087 ** (11.00).00339 ** (4.64) -.00556 ** (-2.75) Only patients who survived to be admitted & with ISS > 15.02880 ** (6.86).01160 * (2.09) -.06370 * (-2.00) * P < 0.05, ** p <.01

39. Relation between treatment at a level 1 or 2 trauma center and death PopulationUnadjusted Beta (t-statistic) AdjustedAdjusted with Instrument All injuries.00512 ** (12.19).00109 ** (4.81) -.00194 ** (-3.07) Injuries with ISS > 15.05859 ** (17.05).00964 ** (2.88) -.04705 * (-2.54) Blunt trauma.00583 ** (12.85).00135 ** (4.66) -.00294 ** (-3.08) Penetrating trauma.00913 ** (8.30).00225 ** (4.16) -.00009 (-0.07) Aged > 55.01087 ** (11.00).00339 ** (4.64) -.00556 ** (-2.75) Only patients who survived to be admitted & with ISS > 15.02880 ** (6.86).01160 * (2.09) -.06370 * (-2.00) * P < 0.05, ** p <.01

40. Unmeasured confounders? Have not fully controlled for case mix? Have not fully controlled for injury severity?  (no physiologic data) The system is intentionally (and effectively) regionalized  the sickest and most complex patients are taken to the highest tier centers Ideally, we would conduct a trial in which we randomize to treatment at a trauma center

41. But we can’t randomize…so… IV

42. Relation between treatment at a level 1 or 2 trauma center and death PopulationUnadjusted Beta (t-statistic) AdjustedAdjusted with Instrument All injuries.00512 ** (12.19).00109 ** (4.81) -.00194 ** (-3.07) Injuries with ISS > 15.05859 ** (17.05).00964 ** (2.88) -.04705 * (-2.54) Blunt trauma.00583 ** (12.85).00135 ** (4.66) -.00294 ** (-3.08) Penetrating trauma.00913 ** (8.30).00225 ** (4.16) -.00009 (-0.07) Aged > 55.01087 ** (11.00).00339 ** (4.64) -.00556 ** (-2.75) Only patients who survived to be admitted & with ISS > 15.02880 ** (6.86).01160 * (2.09) -.06370 * (-2.00) * P < 0.05, ** p <.01

43. Final model examining impact of trauma center on mortality* VariableCoefficientt-statistic Patient Characteristics Age.00005 ** 10.92 Female-.00065 ** -8.50 Medicare (private insurance reference)-.00003-0.19 Medicaid.00043 ** 4.38 Uninsured.00087 ** 7.01 Other payer-.00001-0.13 Injury Characteristics Probability of death.74719 ** 44.32 Intent – self-harm.00828 ** 8.27 Intent – assault-.00218 ** -9.19 Penetrating trauma-.00225 ** -11.93 Hospital Characteristics Not-for-profit (public ownership reference) -.00038-1.49 For-profit-.00005-0.15 Teaching.00129 ** 4.20 Medium hospital (small size reference).003611.83 Large hospital.00106 ** 4.64 Northeast region (West region reference).00043 * 1.99 Midwest region.000080.37 South region.00054 * 2.49 Trauma center-.00194 ** -3.07 * With IV

45. Question 2. Population outcomes for trauma Data Sources (trauma system - supply)  US Census Data –Location of residence at the level of the block group and county  CDC, American Trauma Society, Penn Cartographic Modeling Lab –Trauma center access Data sources (injury death - demand)  National Center for Vital Statistics –Multiple Cause of Death (MCOD) Data

46. Question 2. Population outcomes for trauma Methods  Supply Side – Access to trauma care  Access to level 1/2 trauma center within an hour  Demand Side – Injury Deaths  ICD codes to identify injury death location  Population data to calculate injury death rate Analysis  Examine relation between injury death rates and access to trauma care using poisson distribution

47. Question 2. Population outcomes for trauma Results  Supply  60 minute access to trauma care –84.7% of US residents –46.4% of US counties  Mean time to care = 43 minutes +/- 22  Demand  152,766 injury deaths in 2005  27,964 in counties without access within 60 min  124,802 in counties with access to care within 60 min

50. Counties without access to trauma care within 60 minutes had higher rates of injury death when compared to counties with access to trauma care within 60 minutes (OR 1.24, 95% CI 1.18-1.30) The relative risk of death increased at a rate of 3.4% for each 10 minute increase in time to trauma care (95% CI 2.4% - 4.4%).

51. Q1. Next steps & remaining questions We have a dichotomous outcome variable but are using linear regression… Hard to estimate the strength of the instrument given survey design of NEDS  Is differential distance unrelated to outcome? It would be nice to generate point estimates The direction of effect flips – do you believe it?

52. Q2. Next steps & remaining questions What is the right geographic unit to sum outcomes to?  (we used counties) Should we be targeting counts or rates of death? Should we adjust for injury severity and case mix? (is there systematic variability in severity by geography?)

53. Questions?