Simplified Parsonnet risk scale identifies limits to early patient discharge.

Abstract

Fast-track recovery after coronary artery bypass surgery has influenced patient care positively. Predicting patients who fall off track and require prolonged (> or =7 days) hospitalization remains uncertain. The Parsonnet risk assessment score is effective in predicting length of stay, but is limited by inaccurate subdivision of risk categories. We simplified the Parsonnet risk scale to better identify patients eligible for fast-track recovery. The cases of 604 consecutive patients who underwent isolated coronary artery bypass grafting (CABG) using cardiopulmonary bypass (CPB) were reviewed retrospectively. A rapid recovery protocol emphasizing reduced CPB time, preoperative intra-aortic balloon pump (IABP) criteria, and atrial fibrillation prophylaxis was applied to all patients. The five original divisions of the Parsonnet risk scale were reduced to three risk categories: Low (0-10; Group A), Intermediate (11-20; Group B), High (> 20; Group C). Comparisons of progressive risk categories were analyzed to identify predictive factors associated with fast-track outcomes. The thirty-day operative mortality for the entire group was 3.6%. Three clinical features were identified that distinguished risk progression-female gender, reoperative CABG, and increased age. Additionally, the presence of diabetes (p < 0.05), congestive heart failure (p < 0.01), and peripheral vascular disease (p < 0.001) distinguished Groups A and B, while acute myocardial infarction (p < 0.05) influenced outcomes in Group C. Group A (48%) mean risk score 5.9+/-3.2 was compared to Group B (34%) 14.8+/-2.6, which was further compared to Group C (18%) 26.4+/-2.8. The mean length of stay for Group A (5.3+/-4.1 days) was notably less than Group B (6.1+/-4.7 days; p < 0.05); however, both groups responded favorably to fast-track techniques. Group C did not respond comparably (9.2+/-9.2 vs 6.1+/-4.7 days; p < 0.001) and experienced prolonged recovery. The simplified Parsonnet risk scale did not identify differences in operative mortality and revealed only pneumonia (p < 0.05) and atrial fibrillation (p < 0.01) to be greater in Group C. As risk increased, significantly less revascularization was performed (Group A 3.6+/-1.2 grafts/patient vs Group B 3.3+/-1.2 [p < 0.01]; Group B 3.3+/-1.2 vs Group C 2.5+/-1.0 [p < 0.001]). A simplified Parsonnet risk scale (three categories) is an effective tool in identifying factors limiting fast-track recovery. Low- and intermediate-risk patients represent the majority (82%) and respond well to fast-track methods. High-risk patients (18%) are limited by a greater percentage of female patients, reoperative CABG, and the very elderly, resulting in fast-track failure. Strategies to improve recovery in high-risk patients may include evolving off-pump techniques.