Cecilia Pompili1, Yaron Shargall2, Herbert Decaluwe3, Johnny Moons3, Madhu Chari2, Alessandro Brunelli4. 1. Section of Patient Centred Outcomes Research, Leeds Institute of Cancer and Pathology, University of Leeds, Leeds, UK. 2. Department of Surgery, St. Joseph's Healthcare, McMaster University, Hamilton, CA, USA. 3. Department of Thoracic Surgery, University Hospitals Leuven, Leuven, Belgium. 4. Department of Thoracic Surgery, St. James's University Hospital, Leeds, UK.
Abstract
OBJECTIVES: The objective of this study was to evaluate the performance of 3 thoracic surgery centres using the Eurolung risk models for morbidity and mortality. METHODS: This was a retrospective analysis performed on data collected from 3 academic centres (2014-2016). Seven hundred and twenty-one patients in Centre 1, 857 patients in Centre 2 and 433 patients in Centre 3 who underwent anatomical lung resections were analysed. The Eurolung1 and Eurolung2 models were used to predict risk-adjusted cardiopulmonary morbidity and 30-day mortality rates. Observed and risk-adjusted outcomes were compared within each centre. RESULTS: The observed morbidity of Centre 1 was in line with the predicted morbidity (observed 21.1% vs predicted 22.7%, P = 0.31). Centre 2 performed better than expected (observed morbidity 20.2% vs predicted 26.7%, P < 0.001), whereas the observed morbidity of Centre 3 was higher than the predicted morbidity (observed 41.1% vs predicted 24.3%, P < 0.001). Centre 1 had higher observed mortality when compared with the predicted mortality (3.6% vs 2.1%, P = 0.005), whereas Centre 2 had an observed mortality rate significantly lower than the predicted mortality rate (1.2% vs 2.5%, P = 0.013). Centre 3 had an observed mortality rate in line with the predicted mortality rate (observed 1.4% vs predicted 2.4%, P = 0.17). The observed mortality rates in the patients with major complications were 30.8% in Centre 1 (versus predicted mortality rate 3.8%, P < 0.001), 8.2% in Centre 2 (versus predicted mortality rate 4.1%, P = 0.030) and 9.0% in Centre 3 (versus predicted mortality rate 3.5%, P = 0.014). CONCLUSIONS: The Eurolung models were successfully used as risk-adjusting instruments to internally audit the outcomes of 3 different centres, showing their applicability for future quality improvement initiatives.
OBJECTIVES: The objective of this study was to evaluate the performance of 3 thoracic surgery centres using the Eurolung risk models for morbidity and mortality. METHODS: This was a retrospective analysis performed on data collected from 3 academic centres (2014-2016). Seven hundred and twenty-one patients in Centre 1, 857 patients in Centre 2 and 433 patients in Centre 3 who underwent anatomical lung resections were analysed. The Eurolung1 and Eurolung2 models were used to predict risk-adjusted cardiopulmonary morbidity and 30-day mortality rates. Observed and risk-adjusted outcomes were compared within each centre. RESULTS: The observed morbidity of Centre 1 was in line with the predicted morbidity (observed 21.1% vs predicted 22.7%, P = 0.31). Centre 2 performed better than expected (observed morbidity 20.2% vs predicted 26.7%, P < 0.001), whereas the observed morbidity of Centre 3 was higher than the predicted morbidity (observed 41.1% vs predicted 24.3%, P < 0.001). Centre 1 had higher observed mortality when compared with the predicted mortality (3.6% vs 2.1%, P = 0.005), whereas Centre 2 had an observed mortality rate significantly lower than the predicted mortality rate (1.2% vs 2.5%, P = 0.013). Centre 3 had an observed mortality rate in line with the predicted mortality rate (observed 1.4% vs predicted 2.4%, P = 0.17). The observed mortality rates in the patients with major complications were 30.8% in Centre 1 (versus predicted mortality rate 3.8%, P < 0.001), 8.2% in Centre 2 (versus predicted mortality rate 4.1%, P = 0.030) and 9.0% in Centre 3 (versus predicted mortality rate 3.5%, P = 0.014). CONCLUSIONS: The Eurolung models were successfully used as risk-adjusting instruments to internally audit the outcomes of 3 different centres, showing their applicability for future quality improvement initiatives.
Authors: Marcus Taylor; Syed F Hashmi; Glen P Martin; Michael Shackcloth; Rajesh Shah; Richard Booton; Stuart W Grant Journal: Interact Cardiovasc Thorac Surg Date: 2021-04-08