Race: predictor versus proxy variable? Outcomes after spinal cord injury.

Abstract

To examine the impact of race on acute, rehabilitation, and long-term outcomes after spinal cord injury (SCI). Two case control studies (study 1: acute and rehabilitation outcomes, study 2: long-term outcomes) in which white and nonwhite individuals were matched case for case on multiple demographic, medical, and geographic characteristics with the rationale being that a case-control methodology would increase the internal validity of the design, thereby increasing confidence in the assertion that any between-group differences observed may be specifically attributed to race. Data drawn from the Spinal Cord Injury Model Systems. Institutional practice and general community. Study 1: 187 pairs of individuals, study 2: 158 pairs of matched individuals. Not applicable. Outcome measures assessed included economic (eg, cost of care), treatment-related (eg, length of hospital stay), functional (eg, FIM instrument), and medical (eg, number of medical complications) variables, as well as self-reported life satisfaction, level of handicap, and mental and physical health. In study 1, none of the outcome measures differed significantly across racial groups. Similarly, study 2 failed to indicate significant differences in any of the outcome variables across racial groups, with the exception that nonwhites were at increased risk of greater self-reported handicap in the area of mobility. Power analyses indicated these finding were not merely the result of inadequate power. For the outcomes assessed in studies 1 and 2, race appeared to act primarily as a proxy for other variables (eg, injury severity, age, educational achievement), which in turn may be associated with poor outcome after SCI. Theoretical implications and recommendations are discussed.