The value of patient-reported health status in predicting short-term outcomes after coronary artery bypass graft surgery

BACKGROUND. Risk stratification for comparison of outcomes after coronary artery bypass grafting (CABG) typically includes only clinical measures of risk. Patient-reported health status may be an important independent predictor of short-term health outcomes. OBJECTIVE. To determine whether patient-reported health status, as measured by the Physical and Mental Component Summary scores of the SF-36, predicts in-hospital mortality and prolonged length of stay after CABG, after controlling for other clinical predictors of those outcomes. RESEARCH DESIGN. Prospective cohort study conducted from September 1993 to November 1995. SUBJECTS. One thousand seven hundred seventy-eight adults who underwent isolated CABG for myocardial ischemia. MEASURES. In-hospital mortality and prolonged length of stay (> 14 days). RESULTS. There were 27 deaths and 223 patients with prolonged length of stay in the study sample. A 10-point decrease in the Physical Component Summary (PCS) score increased the odds of in-hospital mortality by 61% (OR, 1.61; 95% CI, 1.04-2.49), independent of established clinical risk factors. Similarly, a 10-point decrease in the PCS score increased the odds of prolonged length of stay by 33% (OR, 1.33; 95% CI, 1.13-1.57). A 10-point decrease in the Mental Component Summary score (MCS) decreased the odds of mortality by 36% (OR, 0.64; 95% CI, 0.43-0.95). CONCLUSIONS. The PCS score is independently and significantly associated with inhospital mortality and prolonged length of stay, after controlling for clinical risk factors. The MCS score is independently and significantly associated only with mortality, though the direction of the effect is unexpected. The result likely reflects a property of the scoring of the MCS and not a finding of clinical substance. Although caution must be taken when interpreting the summary scores, the SF-36 yields information not otherwise captured by clinical data and may be useful in risk stratification for in-hospital mortality and prolonged length of stay after CABG.