Regionalizing Health Care: Volume Standards vs. Risk-Adjusted Mortality Rate

Slide Presentation from the AHRQ 2008 Annual Conference

---

On September 10, 2008, Laurent G. Glance, M.D., made this presentation at the 2008 Annual Conference. Select to access the PowerPoint® presentation (1.1 MB).

---

Slide 1

Regionalizing Health Care: Volume Standards vs. Risk-Adjusted Mortality Rate

• Department of Anesthesiology University of Rochester
• Presented by:
  • Laurent G. Glance, M.D.
    Associate Professor
    Department of Anesthesiology
    University of Rochester
• This project was supported by a grant from the Agency for Healthcare and Quality Research (R01 HS 13617

Slide 2

Team members

• Laurent G Glance, M.D., University of Rochester
• Turner M. Osler, M.D., University of Vermont
• Dana B. Mukamel, Ph.D., University of California, Irvine
• Andrew W. Dick, Ph.D., RAND

Project Officer

• Yen-Pin Chiang, Ph.D.

Slide 3

Scope of the Problem

• Shows a poster of a hand and says, "First, Do No Harm."
• Below that, it says, "To Err Is Human."
• Then it says, "Building a Safer Health System."
• Between 44,000 and 98,000 deaths each year due to medical errors

Slide 4

National Agenda to Improve Patient Safety

• Agency for Healthcare Research and Quality (AHRQ)-sponsored report designated "localizing specific surgeries and procedures to high-volume centers" as a High Priority area for patient safety research.
• Making Health Care Safer: A Critical Analysis of Patient Safety Practices. Evidence Report/Technology Assessment: Number 43. AHRQ Publication No. 01-E058, July 2001. Agency for Healthcare Research and Quality, Rockville, MD.

Slide 5

Interpreting the Volume-Outcome Relationship in the Context of Health Care Quality

Workshop Summary
by Maria Hewitt, for the Committee on Quality of Health Care in America and the National Cancer policy Board

A higher-volume, better-outcome association was observed in three-quarters of the studies reviewed. Volume is, however, an imprecise indicator of quality. Some low-volume providers have excellent outcomes, and, conversely, some high-volume providers have poor outcomes. Volume per se does not lead to good outcomes in health care; it is instead a proxy measure for other factors that affect care. With few exceptions, however, the literature does not shed light on the structures or processes of care that underlie the apparent relationship.

Slide 6

Special Article from The New England Journal of Medicine

• Hospital Volume and Surgical Mortality in the U.S., by
  John D. Birkmeyer, M.D., Andrea E. Siewers, M.P.H., Emily V.A. Finlayson, M.D., Therese A. Stukel, Ph.D., F. Lee Lucas, Ph.D., Ida Batista, B.A., H. Gilbert Welch, M.D., M.P.H., and David E. Wennberg, M.D., M.P.H.
• Conclusions: In the absence of other information about the quality of surgery at the hospitals near them, Medicare patients undergoing selected cardiovascular or cancer procedures can significantly reduce their risk of operative death by selecting a high-volume hospital. (N Engl J Med 2002;346:1128-37.)

Slide 7

Selective Referral to High-Volume Hospitals; Estimating Potentially Avoidable Deaths

From a article in JAMA, March 1 2000—Vol 283, No. 9,
by:
R. Adams Dudley, MD, MBA
Kirsten L. Johansen, MD
Richard Brand, PhD
Deborah J. Rennie
Arnold Milstein, MD, MPH

Table 3. Calculation of Potentially Avoidable Deaths

Condition	No. Admitted to LVHs (% of Total Admissions for This Condition)	Deaths in LVHs	Point Estimate of Excess Deaths in LVHs (95% CI)	Excess Deaths in LVHs as Percentage of All Deaths in LVHs (95% CI)	No. of HVHs/LVHs in California in 1997
Coronary artery bypass surgery	21 452 (66)	952	258 (124-372)	27 (13-39)	19/107
Lower extremity arterial bypass surgery	4054 (39)	194	37 (26-48)	19 (13-25)	78/245
Heart Transplantation	21 (8)	0	0	0	9/4
Pediatric heart surgery	468 (21)	26	7 (4-10)	27 (15-38)	8/33
Coronary angioplasty	16 070 (36)	327	80 (9-122)	24 (3-37)	44/99
Elective abdominal aortic aneurysm repair	1837 (64)	107	40 (16-58)	37 (15-54)	26/269
Carotid endarterectomy	7476 (63)	73	16 (8-23)	22 (11-32)	32/264
Cerebral aneurysm surgery	818 (73)	83	36 (26-45)	43 (31-54)	7/145
Esophageal cancer surgery	130 (77)	9	7 (4-7)	78 (44-78)	3/77
Pancreatic cancer surgery	271 (57)	27	20 (8-22)	74 (30-85)	15/138
HIV/AIDS	5709 (37)	475	101 (79-123)	21 (17-26)	36/327
Total	58 306 (48)	2273	602 (304-830)	26 (13-37)	...

Note: LVHs = low-volume hospitals; HVHs = high-volume hospitals.

Slide 8

Hypotheses

• Selective Referral: Selectively referring high-risk surgery patients to high-quality centers will lead to better population outcomes than selectively referring patients to high-volume centers.
• Selective Avoidance: Diverting high-risk patients from low quality centers will lead to better population outcomes than diverting patients from low-volume centers.

Slide 9

Data

• Healthcare Cost and Utilization Project (HCUP) California State Inpatient Databases (SID) (1998-2000)
• Administrative data (ICD-9-CM codes):
  • 30 diagnoses
  • 21 procedures
  • Present on Admission (POA) indicator
• Study Populations:
  • Coronary Artery Bypass Grafting (CABG)
  • Percutaneous coronary intervention (PCI)
  • Abdominal aoratic aneurysm (AAA) surgery

Slide 10

Model Development

Shows a sample equation

• Random-Intercept model
• Demographics:
  • Age, gender, transfer status, admission type (elective vs. non-elective)
• Comorbidities:
  • Disease Staging
  • Elixhauser Comorbidity Algorithm

Slide 11

Hospital "Quality"

Among Build Research Infrastructure Capacity (BRIC) scientists:

• Hospital intercept term

Slide 12

Identification of High-Volume and Low-Volume Centers

• High-Volume based on Leapfrog Criteria:
  • AAA > 50 cases/yr
  • CABG > 450 cases/yr
  • PCI > 400 cases/yr
• Low-Volume:
  • Lower volume quartile

Slide 13

Estimating Impact of Regionalization

• Added binary variable to base model to indicate whether a patient was treated at a high-volume center.
• Simulated mortality rate:
  • Mortality rate for patients diverted to high-volume centers.
  • Estimated Observed mortality rate for patients already treated at high-volume centers.

Slide 14

Shows a cartoon with two men at a blackboard.

There are equations to the left and right, and one man is pointing to a phrase that says, "Then a Miracle Occurs," and the other man is saying, "I think you should be more explicit in step two."

Slide 15

Volume-Outcome Association

• Hospital volume is NOT a good proxy for Hospital Quality

Slide 16

Coronary Artery Bypass Surgery

• Line graphs show that high-quality hospitals have a lower risk-adjusted mortality rate than low-quality hospitals.

Slide 17

Impact of Regionalization

Table 2. Regionalization based on a strategy of selective referral and avoidance

		Baseline			Leapfrog				High quality
		Mortality (%)	Patients	Hospitals	Mortality (%)	Relative risk reduction (%)	Diverted patients (%)	Hospitals closed (%)	Mortality (%)	Relative risk reduction	Diverted patients (%)	Hospitals closed (%)
Selective referral	AAA	12.55	8855	301	11.51	8.3	93.7	99.0	5.87	53	96.6	99.3
	CABG	3.02	84 586	123	2.42	20	69.6	91.1	1.58	48	82.9	91.9
	PCI	1.56	149 375	162	1.53	1.9	29.6	67.3	0.72	54	95.6	97.5
Selective avoidance	AAA	12.55	8855	301	12.36	1.5	3.2	28.2	12.33	1.8	3.1	0.7
	CABG	3.02	84 656	123	2.94	2.6	7.1	26	2.87	5.0	6.7	8.1
	PCI	1.56	149 375	162	1.56	0	0.20	25.3	1.47	5.8	8.7	5.6

Notes: AAA, abdominal aoratic aneurysm surgery; CABG, coronary bypass surgery; PCI, percutaneous coronary intervention.
P ≤ 0.05 compared to baseline.

Slide 18

Findings

• Selective Referral:
  • High-Volume Centers: 0-20% mortality reduction & 70-99% hospital closure
  • High-Quality Centers: 50% mortality reduction & 90-99% hospital closure
• Selective Avoidance:
  • Low-Volume Centers: 0-2.5% reduction in mortality & 25% hospital closure
  • Low-Quality Centers: 2-5% mortality reduction & 1-8% hospital closure

Slide 19

Policy Implications

• Hospital Volume is a POOR Quality Indicator & should not be used as the basis for selective referral or selective avoidance.
• Selective Referral to High-Quality Centers is NOT PRACTICAL.
• Selective Avoidance of Low-Quality Centers may achieve modest reductions in mortality.
• Consider Improving Overall Hospital Quality.

Slide 20

Quality Improvement based on Feedback of Risk-Adjusted Outcomes

• National Surgical Quality Improvement Program (NSQIP)
• Northern New England (NNE)

Slide 21

NSQIP: Khuri Arch Surgery 2002

• Line graphs show:
  • 27% decrease in mortality
  • 45% decrease in morbidity
  • No change in casemix

Slide 22

NNE Cardiovascular Study by O'Connor GT. JAMA 1996

• Line graphs compare expected versus observed mortality and indicate that mortality drops after intervention.

Slide 23

Current Project

Current Project

Slide 24

SMARTT—Survival Measurement and Reporting Trial for Trauma

• Project Officer
• Michael Handrigan, PhD
• Investigators: Laurent G. Glance et al
• Funded by the Agency for Healthcare Research and Quality (R01 HS 16737)

Slide 25

Hypothesis

• Providing trauma and non-trauma centers with information on their risk-adjusted outcomes will lead to improved outcomes.

Slide 26

Hospital Odds Based on All Trauma Cases

Graph concludes that the odds of a trauma patient dying inyour hospital are 0.49 compared to the average hospital (95% confidence interval: 0.33, 0.74). The likelihood that your hospital is a high-quality hospital (OR <1) is 100%. The likelihood that your hospital is a very-high-quality hospital (OR <0.8) is 99%.

Slide 27

Hospital Odds Ratio Based on Mechanism of Trauma

• Trauma Center Quality 2006—Blunt Trauma concludes that the odds of a blunt trauma patient dying in your hospital are 0.80 compared with an average hospital.
• Trauma Center Quality 2006—GSW Trauma concludes that the odds of a gunshot wound trauma patient dying in your hospital are 0.49 compared with an average hospital.