M. Melis1,2, A. Masi1,2, I. Hatzaras1,2, G. Ballantyne1,2, A. Pinna1,2, A. Rosman1, D. Neihaus1, S. Cohen1,2, J. K. Saunders1,2, E. Newman1,2, R. S. Berman1,2, H. Pachter1,2 1New York Harbor Healthcare System VAMC,New York, NY, USA 2New York University School Of Medicine,New York, NY, USA
Background: The VASQIP (Veteran Affairs Surgical Quality Improvement Program) risk calculator may predict post-operative morbidity and mortality (M&M) of non-emergent procedures based on pre-existing patient conditions and magnitude of surgery. The Surgical Apgar Score (SAS) may predict M&M based on three simple intra-operative measurements. Unlike VASQIP, which is calculated pre-operatively, SAS is entirely dependent on intra-operative variables and may be influenced by the performance of a surgical team. We hypothesized that SAS could potentially be used as a measure of intra-operative performance across patients with similar VASQIP risk undergoing major surgery of the alimentary tract.
Methods: We categorized patients undergoing major surgery of the alimentary tract at the NY Harbor VAMC in 3 risk categories (low, average, high risk) according to their VASQIP score. We then evaluated whether SAS could predict MM within each VASQIP risk category. Multivariate logistic regression was used to assess the effect of SAS on the study’s two end-points: postoperative morbidity and mortality.
Results: From October 2006 to July 2011, 560 patients underwent major surgery of the alimentary tract. There were 545 (97.3%) men and 15 (2.7%) women. Average age was 68.4 ± 12 years. According to the pre-operative VASQIP estimate for risk of complications, the following study groups were identified: low risk (n=19), average risk (n=156), high risk (n=385). Of note, all patients with SAS <4 were clustered in the VASQIP high-risk group. Overall morbidity was 37.5% and overall mortality was 5.7%. M&M were 56.2% and 25% for SAS <4 and 23.1% and 0% for SAS >8, respectively. In high risk patients SAS was strongly correlated with morbidity (56.2% vs. 32%, p<0.05) and mortality (25% vs. 0%, p<0.05). Within the high-risk group low SAS accurately predicted M&M: [for SAS <6, OR= 1.86, 95% CI (1.18, 2.92), p= 0.007]. Also, within the high risk group, a high SAS was “protective” toward M&M: [for SAS >8, OR=0.47, 95%CI (0.28, 0.81), p=0.006]. Across low and average risk groups SAS did not influence M&M: however, lack of correlation in those two groups is likely secondary to a type II error.
Conclusions: SAS is strongly associated with increased morbidity and mortality in patients classified as high risk for complications by VASQIP. In fact, SAS is a sensitive tool that can differentiate patients at highest risk of postoperative M&M, even within the highest preoperative risk assessment by VASQUIP. In this way SAS offers supplemental information that can direct patient disposition and further care. Additional studies are needed to validate this concept and establish SAS as a surgical quality measurement.