Original Article of Interventional Pulmonology Corner
Nomogram to predict postoperative PR in patients undergoing CT-guided transthoracic lung biopsy
Abstract
Background: Pleural reaction (PR) frequently occurs following computed tomography-guided transthoracic needle biopsy (CT-TNB). The purpose of this study was to establish a predictive model for PR following CT-TNB.
Methods: In this study, a total of 436 patients who underwent CT-TNB between June 2016 and December 2017 at a tertiary hospital were consecutively included. Patient demographics, lesion features, laboratory tests, and biopsy parameters were collected. The least absolute shrinkage and selection operator (LASSO) regression and multivariate logistic regression analyses were performed to establish a prediction model for post-CT-TNB PR, presented by a nomogram. Discrimination and calibration were assessed. For internal validation, a bootstrap resampling method was applied, and decision curve analysis (DCA) was used to evaluate its clinical utility.
Results: PR occurred in 7.8% (34/436) of patients. Four non-zero coefficient variables (gender, age, lesion location, and puncture position) were filtered by LASSO regression analysis and were used to establish a predictive model. The area under the curve in the derivation and validation was 0.840 (95% CI, 0.767–0.913) and 0.841 (95% CI, 0.769–0.912), respectively. The model was well-calibrated (P>0.05), and DCA indicated clinical efficacy.
Conclusions: In this study, we established a nomogram, including as parameters gender, age, lesion location, and puncture position, which may have great significance for individualized prediction of post-CTTNB PR.
Methods: In this study, a total of 436 patients who underwent CT-TNB between June 2016 and December 2017 at a tertiary hospital were consecutively included. Patient demographics, lesion features, laboratory tests, and biopsy parameters were collected. The least absolute shrinkage and selection operator (LASSO) regression and multivariate logistic regression analyses were performed to establish a prediction model for post-CT-TNB PR, presented by a nomogram. Discrimination and calibration were assessed. For internal validation, a bootstrap resampling method was applied, and decision curve analysis (DCA) was used to evaluate its clinical utility.
Results: PR occurred in 7.8% (34/436) of patients. Four non-zero coefficient variables (gender, age, lesion location, and puncture position) were filtered by LASSO regression analysis and were used to establish a predictive model. The area under the curve in the derivation and validation was 0.840 (95% CI, 0.767–0.913) and 0.841 (95% CI, 0.769–0.912), respectively. The model was well-calibrated (P>0.05), and DCA indicated clinical efficacy.
Conclusions: In this study, we established a nomogram, including as parameters gender, age, lesion location, and puncture position, which may have great significance for individualized prediction of post-CTTNB PR.
