Start Your Search
MINI 19 - Surgical Topics in Localized NSCLC (ID 138)
- Event: WCLC 2015
- Type: Mini Oral
- Track: Treatment of Localized Disease - NSCLC
- Presentations: 1
MINI19.08 - Validation of a Surgical Predictive Score for 90 Day Mortality in Lung Cancer and Comparison with Thoracoscore (ID 2754)
16:45 - 18:15 | Author(s): H.A. Powell
Current British Thoracic Society (BTS) guidelines advocate the use of a global risk prediction score such as Thoracoscore to estimate the risk of death prior to radical surgical management in those with non-small cell lung cancer (NSCLC). A recent publication by Powell et al(1) used the National Lung Cancer Audit (NLCA) linked to Hospital Episode Statistics (HES) to produce a score to predict 90 day mortality. The aim of this study is to validate this score, henceforth called the NLCA score, and compare its performance with Thoracoscore.
We identified data on all patients in the NLCA who received curative surgery for NSCLC between 2010 and 2012. We calculated the proportion that died in hospital and within 90 days of surgery. Each person was given a score based on the coefficients and constants in the NLCA score and Thoracoscore. The discriminatory power of both scores was assessed by a receiver operating characteristic (ROC) and an area under the curve (AUC) calculation.
We identified 2858 patients for whom we had complete data to form our validation cohort. The 90 day mortality was 5%. We generated ROC curves to assess the discrimination of the NLCA score in predicting 90 day mortality and to test the ability of Thoracoscore to predict in-hospital mortality. Area under the ROC curve was 0.68 and 0.60 respectively. We performed a post hoc analysis using data from the NLCA on all 15554 patients who underwent curative surgery for NSCLC between 2004 and 2012 to derive summary tables for 90 day mortality, stratified by procedure type, age and performance status (table 1).
These results suggest that although the NLCA score performs slightly better than Thoracoscore neither performs well enough to be advocated for routine use to risk stratify patients prior to lung cancer surgery. It may be that the addition of physiological parameters to demographic and procedural data or use of physiological measurements alone would better predict mortality; however this would form the basis of a further project. In the interim we advocate the use of our summary tables that serve to provide clinicians and patients the real-life range of mortality according to performance status and age for both lobectomy and pneumonectomy. 1. Powell HA, Tata LJ, Baldwin DR, Stanley RA, Khakwani A, Hubbard RB. Early mortality after surgical resection for lung cancer: an analysis of the English National Lung cancer audit. Thorax. 2013;68(9):826-34. Figure 1
Only Active Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login or select "Add to Cart" and proceed to checkout.