MQMS: Diabetes, 19922001

1
MQMS Diabetes, 1992-2001

• Sylvia Kuo, Ph.D. (MPR)
• Barbara Fleming, M.D., Ph.D. (VA)
• Neil Gittings, M.A. (CMS)
• Lein Han, Ph.D. (CMS)
• Linda Geiss, M.A. (CDC)
• Michael Engelgau, M.D., M.S. (CDC)
• Sheila Roman, M.D., M.P.H. (CMS)
• Presented by
• Angela Merrill, PhD (MPR)
• June 2004

2
Importance of Monitoring Diabetes in Medicare

• Rising economic cost and prevalence of diabetes
• 98 billion in 1997 to 132 billion in 2002
• Diabetes can lead to severe and costly outcomes
• Healthy People 2010 has targets specific to
  diabetes
• Disproportionate cost of diabetes borne by
  Medicare

3
Study Questions

• How has quality of diabetes care in Medicare
  changed over time?
• Is there evidence of subgroup disparities?
• What areas of concern might there be for
  policymakers?

4
Data

• Annual cross-sections from 1992-2001
• 5 Standard Analytic Files
• Inclusion criteria
• Full year Medicare FFS participant
• Principal or secondary diagnosis of diabetes on
  at least 1 hospitalization/ED visit OR 2
  ambulatory encounters
• Exclude if gestational diabetes or ESRD

5
Process and Outcome Measures

• Preventive care practices
• HbA1c test
• Lipid profiling
• Eye exam
• Self-monitoring of glucose
• Short-term acute complications
• Hospitalization for ketoacidosis
• ED visit for metabolic crisis
• Long-term complications
• Nephropathy
• Lower limb amputation
• Blindness retinopathy
• Hospitalization for AMI, heart failure, stroke

6
National Trends in Preventive Service Receipt,
1992-2001
Per 100 Medicare Beneficiaries with Diabetes
95 confidence intervals within /- 0.37 per 100
for all measures and years.
7
National Trends in Short-term Complications,
1992-2001
Per 1,000 Medicare Beneficiaries with Diabetes
95 confidence intervals within /- 0.55 per
1,000 for all measures and years.
8
National Trends in Long-term Complications,
1992-2001
Per 1,000 Medicare Beneficiaries with Diabetes
95 confidence intervals within /- 0.62 per
1,000 for all measures and years.
9
National Trends in Long-term Complications,
1992-2001
Per 1,000 Medicare Beneficiaries with Diabetes
95 confidence intervals within /- 0.62 per
1,000 for all measures and years.
10
National Trends in Long-term Complications,
1992-2001
Per 1,000 Medicare Beneficiaries with Diabetes
95 confidence intervals within /- 1.12 per
1,000 for all measures and years.
11
Disparities Among Patient Subgroups

• Some subgroups have
• Lower preventive service rates
• Higher complication rates
• These subgroups are
• Non-whites (vs. whites)
• Dual eligibles (vs. non-dually eligible)
• Oldest of the old 85
• Under 65

12
Example Disparities in Preventive Service Rates
by Race
Difference between subgroup and all other
subgroups statistically significant at the 5
level.
13
Comparison with Healthy People 2010 Goals
14
Summary

• Improvements
• ? preventive care practices
• ? short-term complications
• ? amputations, cardiovascular conditions
• Challenges
• ? nephropathy, eye complications
• Disparities in care for nonwhite, dual eligible,
  lt65, 85
• Progress towards Healthy People 2010 goals

15
Limitations

• Subject to changes in billing and coding
  practices
• Underestimate overall prevalence
• May reflect improved detection of complications,
  not increasing prevalence

16
Implications for Policy and Clinical Practice

• MQMS identifies areas of improvements in diabetes
  quality of care
• Identifies where opportunities remain
• Stimulates research on disparities for vulnerable
  populations