Race and outcomes after percutaneous coronary intervention: Insights from the Blue Cross Blue Shield of Michigan Cardiovascular Consortium
Authors' conclusion
Black patients had a higher risk of 90-day readmission and cumulative mortality following PCI compared with White patients. Associations were mediated by dual eligibility, community economic well-being, and traditional cardiovascular risk factors.
Study context
- Population
- Black and White patients undergoing percutaneous coronary intervention (PCI) at 48 Michigan hospitals, January 2013–March 2018, Medicare fee-for-service10.28% Black, 89.72% White
- Participants
- 29,317 participants
- Exposure
- Black vs White patients after PCI
- Outcome
- 90-day readmission and postdischarge all-cause mortality
Methodological detailspropensity-score matched, 48 Michigan hospitals, registry linked to Medicare claims, mediation analysis
- Propensity-score matched
- 48 Michigan hospitals
- Registry linked to Medicare claims
- Mediation analysis
Main finding
As reported in the abstract
Black vs WhiteEstimate (95% CI)
90-day readmission; adjusted for age and gender
adjusted OR 1.62(1.32–2.00)
All-cause mortality; adjusted for age and gender
adjusted HR 1.45(1.30–1.61)
Observational data: associations, not effects.
Further findings
Proportion of association mediated (PM)
- Dual eligibility: readmission
- 11.0%
- Dual eligibility: mortality
- 21.1%
- Dual eligibility + community economic well-being: readmission
- 22.3%
- Dual eligibility + community economic well-being: mortality
- 43.0%
- Dual eligibility + community economic well-being + baseline clinical characteristics: readmission
- 45.0%
- Dual eligibility + community economic well-being + baseline clinical characteristics: mortality
- 87.8%
- Significantly mediated
Post-PCI in-hospital outcomes
- Minimal differences between groups
Original publication
Race and outcomes after percutaneous coronary intervention: Insights from the Blue Cross Blue Shield of Michigan Cardiovascular Consortium. Am Heart J. Published 8 October 2022. doi:10.
Summarised from the published abstract; every number is checked against it