Development and validation of a risk score for predicting death after pneumonectomy

Comparison of standard and penalized logistic regression in risk model development

Objective

Regression models are ubiquitous in thoracic surgical research. We aimed to compare the value of standard logistic regression with the more complex but increasingly used penalized regression models using a recently published risk model as an example.

Methods

Using a standardized data set of clinical T1-3N0 esophageal cancer patients, we created models to predict the likelihood of unexpected pathologic nodal disease after surgical resection. Models were fitted using standard logistic regression or penalized regression (ridge, lasso, elastic net, and adaptive lasso). We compared the model performance (Brier score, calibration slope, C statistic, and overfitting) of standard regression with penalized regression models.

Results

Among 3206 patients with clinical T1-3N0 esophageal cancer, 668 (22%) had unexpected pathologic nodal disease. Of the 15 candidate variables considered in the models, the key predictors of nodal disease included clinical tumor stage, tumor size, grade, and presence of lymphovascular invasion. The standard regression model and all 4 penalized logistic regression models had virtually identical performance with Brier score ranging from 0.138 to 0.141, concordance index ranging from 0.775 to 0.788, and calibration slope from 0.965 to 1.05.

Conclusions

For predictive modeling in surgical outcomes research, when the data set is large and the outcome of interest is relatively frequent, standard regression models and the more complicated penalized models are very likely to have similar predictive performance. The choice of statistical methods for risk model development should be on the basis of the nature of the data at hand and good statistical practice, rather than the novelty or complexity of statistical models.

Pneumonectomy for primary lung cancer: contemporary outcomes, risk factors and model validation

OBJECTIVES

Despite the increased rate of adverse outcomes compared to lobectomy, for selected patients with lung cancer, pneumonectomy is considered the optimal treatment option. The objective of this study was to identify risk factors for mortality in patients undergoing pneumonectomy for primary lung cancer.

METHODS

Data from all patients undergoing pneumonectomy for primary lung cancer at 2 large thoracic surgical centres between 2012 and 2018 were analysed. Multivariable logistic and Cox regression analyses were used to identify risk factors associated with 90-day and 1-year mortality and reduced long-term survival, respectively.

RESULTS

The study included 256 patients. The mean age was 65.2 (standard deviation 9.4) years. In-hospital, 90-day and 1-year mortality were 6.3% (n = 16), 9.8% (n = 25) and 28.1% (n = 72), respectively. The median follow-up time was 31.5 months (interquartile range 9-58 months). Patients who underwent neoadjuvant therapy had a significantly increased risk of 90-day [odds ratio 6.451, 95% confidence interval (CI) 1.867-22.291, P = 0.003] and 1-year mortality (odds ratio 2.454, 95% CI 1.079-7.185, P = 0.044). Higher Performance Status score was associated with higher 1-year mortality (odds ratio 2.055, 95% CI 1.248-3.386, P = 0.005) and reduced overall survival (hazard ratio 1.449, 95% CI 1.086-1.934, P = 0.012). Advanced (stage III/IV) disease was associated with reduced overall survival (hazard ratio 1.433, 95% CI 1.019-2.016, P = 0.039). Validation of a pneumonectomy-specific risk model demonstrated inadequate model performance (area under the curve 0.54).

CONCLUSIONS

Pneumonectomy remains associated with a high rate of perioperative mortality. Neoadjuvant chemoradiotherapy, Performance Status score and advanced disease emerged as the key variables associated with adverse outcomes after pneumonectomy in our cohort.

A systematic review of risk prediction models for perioperative mortality after thoracic surgery

OBJECTIVES

Guidelines advocate that patients being considered for thoracic surgery should undergo a comprehensive preoperative risk assessment. Multiple risk prediction models to estimate the risk of mortality after thoracic surgery have been developed, but their quality and performance has not been reviewed in a systematic way. The objective was to systematically review these models and critically appraise their performance.

METHODS

The Cochrane Library and the MEDLINE database were searched for articles published between 1990 and 2019. Studies that developed or validated a model predicting perioperative mortality after thoracic surgery were included. Data were extracted based on the checklist for critical appraisal and data extraction for systematic reviews of prediction modelling studies.

RESULTS

A total of 31 studies describing 22 different risk prediction models were identified. There were 20 models developed specifically for thoracic surgery with two developed in other surgical specialties. A total of 57 different predictors were included across the identified models. Age, sex and pneumonectomy were the most frequently included predictors in 19, 13 and 11 models, respectively. Model performance based on either discrimination or calibration was inadequate for all externally validated models. The most recent data included in validation studies were from 2018. Risk of bias (assessed using Prediction model Risk Of Bias ASsessment Tool) was high for all except two models.

CONCLUSIONS

Despite multiple risk prediction models being developed to predict perioperative mortality after thoracic surgery, none could be described as appropriate for contemporary thoracic surgery. Contemporary validation of available models or new model development is required to ensure that appropriate estimates of operative risk are available for contemporary thoracic surgical practice.

Development of a nomogram for predicting the operative mortality of patients who underwent pneumonectomy for lung cancer: a population-based analysis

Background

Although many studies have reported that patients have undergone entire lung removal for lung cancer along with high operative mortality, the trends in the incidence and associated risk factors for operative death have not been explored in a national population-based study. In addition, a clinical decision-making nomogram for predicting postpneumonectomy mortality remains lacking.

Methods

A total of 10,337 patients diagnosed with lung cancer who underwent pneumonectomy between 1998 and 2016 were retrieved from the Surveillance, Epidemiology, and End Results (SEER) cancer registry. Multivariate logistic regression analysis was used to identify risk factors for predicting operative mortality. Thereafter, these independent predictors were integrated into a nomogram, and bootstrap validation was applied to assess the discrimination and calibration. Additionally, decision curve analysis (DCA) was used to calculate the net benefit of this forecast model.

Results

The overall postpneumonectomy mortality between 1998 and 2016 was 10.3%, including a 30-day mortality of 4.2%; however, there were statistically significant decreases in the operative death rates from 8.8% in 1998 to 6.7% in 2016 (P=0.009). Higher operative mortality was associated with advanced patients (P<0.001), male sex (P<0.001), right-sided pneumonectomy (P<0.001), squamous cell carcinoma (SCC) (P=0.008), number of positive lymph nodes (npLNs) 5 or greater (P=0.010), and distant metastasis (P<0.001). However, induction radiotherapy (RT) was a protective factor (P<0.001). The nomogram integrating all of the above independent predictors was well calibrated and had a relatively good discriminative ability, with a C-statistic of 0.687 and an area under the receiver operating characteristic (ROC) curve (AUC) of 0.682; moreover, DCA demonstrated that our model was clinically useful.

Conclusions

If pneumonectomy was considered inevitable, clinical decision-making based on this simple but efficient predictive nomogram could help minimize the risk of operative death and maximize the survival benefit.

Ninety-Day Mortality: Redefining the Perioperative Period After Lung Resection

Operative mortality is an important outcome for patients, surgeons, healthcare institutions, and policy makers. Although measures of perioperative mortality have conventionally been limited to in-hospital and 30-day mortality (or a composite endpoint combining both), there is a large body of evidence emerging to support the extension of the perioperative period after lung resection to a minimum of 90 days after surgery. Several large-volume studies from centers across the world have reported that 90-day mortality after lung resection is double 30-day mortality. Hence, true perioperative mortality after lung resection is likely to be significantly higher than what is currently reported. In the contemporary era, where new treatment modalities such as stereotactic ablative body radiotherapy are emerging as viable nonsurgical alternatives for the treatment of lung cancer, accurate estimation of perioperative risk and reliable reporting of perioperative mortality are of particular importance. It is likely that shifting the discussion from 30-day to 90-day mortality will lead to altered decision making, particularly for specific patient subgroups at an increased risk of 90-day mortality. We believe that 90-day mortality should be adopted as the standard measure of perioperative mortality after lung resection and that strategies to reduce the risk of mortality within 90 days of surgery should be investigated.