Virtual Library
Start Your Search
M. Meza Guzman
Author of
-
+
P1.06 - Poster Session/ Screening and Early Detection (ID 218)
- Event: WCLC 2015
- Type: Poster
- Track: Screening and Early Detection
- Presentations: 1
- Moderators:
- Coordinates: 9/07/2015, 09:30 - 17:00, Exhibit Hall (Hall B+C)
-
+
P1.06-018 - Can 30-Mortality after Lung Cancer Resection Be Used as an Individual Surgeon Quality Outcome Internationally? National Data from the UK (ID 2866)
09:30 - 09:30 | Author(s): M. Meza Guzman
- Abstract
Background:
Internationally, one of the most commonly reported quality outcome in surgery for lung cancer is 30 day mortality. However, is difficult to know what constitutes unacceptably high mortality or unacceptable variation between surgeons. In October 2014 national data was released from the Society for Cardiothoracic Surgery (SCTS) in the United Kingdom (UK) on hospital and individual surgeon volume performance for lung cancer resection in the UK. The implicit assumption is benchmarking of the performance. The aim of this study is to report on the impact of individual surgeon volume in relation to each death associated with the an average 30-day mortality rate of 2.2% using national data driven performance control limits (i.e. funnel plots), and determine the applicability on surgeon performance internationally.
Methods:
Data released by the SCTS were downloaded, complied and analysed. Each step change for individual mortality was calculated, and alert limits modelled using current UK national standard of the upper 99% binomial confidence limit.
Results:
Data from 29 units were published with the annual volume of 125 surgeons for 2012. Data from 6 surgeons were excluded for no lung resections performed. In the remaining 118 surgeons, the mean (SD) annual lung resection volume for cancer was 42 (27). A total of 25% of surgeons performed 18 resections (or less) per year. For 50% of surgeons undertaking 40 resections (or less) each death represents at least 2.5% (0 to 13%) of their annual work load. Using a 99% binomial confidence limit at 50 cases, the upper alert is 16%. Therefore for the majority of surgeons, a mortality rate of 15% which is 7.5 fold higher than average would not trigger the conventional national alert limits.
Conclusion:
Based on UK national data, lung cancer resection volumes for individual surgeons are low and for the majority even a single death (which could be due to chance), affects the overall mortality rate much more, carries a disproportionately high weighting and may encourage risk adverse behaviour whilst simultaneously failing to detect 7.5 fold increased mortality rates using conventional national limits. Such data driven limits would also not be applicable on an international level basis unless individual surgeon volume is higher than 100 resections per year, a level that was not achieved by most UK surgeons.
