Impact of ethnic variation and residential segregation on long-term survival following myocardial perfusion SPECT

Abstract

Background: Ethnic characteristics of a neighborhood may impact upon all-cause mortality (ACM). It is not known whether this consideration remains a risk modifier among those being evaluated for CAD.

Methods: 6,477 pts (60 ± 13 years, male 38%) residing in NYC with normal or abnormal stress SPECT studies were assessed for ACM during a mean follow-up of 9 ± 3.8 years. Baseline CAD risk factors and ethnic characteristics of patient neighborhoods were considered. Zip-codes with >70% of one ethnicity was considered to be predominant of that ethnicity.

Results: There were 573 (20%) Hispanics (HS), 765 (27%) African-Americans (AA), and 250 (30%) Caucasians (CC) residing in areas >70% of their own ethnicity. Compared to CC, the risk for ACM was lower in HS (hazard ratio (HR) 0.68, 95% CI 0.57-0.8, P < .0001) and similar among AA (HR 1.1, 95% CI 0.95-1.41, P = .2). Among HS, there was a lower ACM among those residing in HS areas compared to those residing in a non-HS areas (HR 0.7 95% CI 0.56-0.9, P = .03) despite a lower median household income ($27,838 ± 3,328 vs $37,751 ± 17,036; P < .0001). This survival difference was not seen in CC and AA.

Conclusion: Among patients referred for nuclear SPECT studies for suspected CAD, HS ethnicity was an independent predictor of a favorable prognosis. Among HS, the ethnic characteristic of patients' neighborhoods was an independent predictor of ACM. These results imply that ethnic social support is a potentially powerful modifier of patient outcomes among certain patient groups.