Risk prediction model for respiratory complications after lung resection
Autor: | Guijarro-Jorge Ricardo, MAIRA BES-RASTROLLO, Marc Gimenez-Milà, Pablo Monedero |
---|---|
Rok vydání: | 2016 |
Předmět: |
Male
medicine.medical_specialty Clinical prediction rule 030204 cardiovascular system & hematology Risk Assessment Decision Support Techniques 03 medical and health sciences 0302 clinical medicine Predictive Value of Tests Risk Factors Forced Expiratory Volume medicine Humans Lung surgery Pneumonectomy Lung Aged Retrospective Studies Chi-Square Distribution business.industry Process Assessment Health Care Smoking Age Factors Reproducibility of Results Retrospective cohort study Middle Aged Respiration Disorders Surgery Logistic Models Treatment Outcome Anesthesiology and Pain Medicine ROC Curve 030228 respiratory system Spain Cardiothoracic surgery Area Under Curve Predictive value of tests Multivariate Analysis Linear Models Observational study Risk assessment business Chi-squared distribution |
Zdroj: | European Journal of Anaesthesiology. 33:326-333 |
ISSN: | 0265-0215 |
Popis: | Patients undergoing lung surgery are at risk of postoperative pulmonary complications (PPCs). Identifying those patients is important to optimise individual perioperative management. The Clinical Prediction Rule for Pulmonary Complications (CPRPCs) after thoracic surgery, developed by the Memorial Sloan-Kettering Cancer Center, might be an ideal predictor. The hypothesis was that CPRPC performs well for the prediction of PPCs.The aim of our study was to provide the external validation of the CPRPC after lung resection for primary tumours, before universal acceptance. In case of poor discrimination, we planned, as a second objective, to derive a new predictive index for PPCs.Retrospective, observational multicentre study.A total of 559 adult consecutive patients who underwent pulmonary resection. Inclusion criteria were adult patients (aged over 17 years).Thirteen Spanish hospitals during the first half of 2011.A record of the PPCs defined, as in the original publication, as the presence of any of the following events: atelectasis; pneumonia; pulmonary embolism; respiratory failure; and need for supplemental oxygen at hospital discharge.The performance of the CPRPC was determined in order to examine its ability to discriminate and calibrate the presence of PPCs.The study included 559 patients, of whom 75 (11.6%) suffered PPCs. The CPRPC did not show enough discriminatory power for our cohort area under the receiver operating characteristic (ROC) curve 0.47 (95% confidence interval 0.37 to 0.57)]. After a fitting step by stepwise multivariate logistic regression, we identified three main predictors of PPCs: age; smoking status; and predicted postoperative forced expiratory volume in 1 s. Combining them into a simple risk score, we were able to obtain an area under the ROC curve of 0.74 (95% confidence interval 0.68 to 0.79).In this external validation, the CPRPC performed poorly despite its simplicity. The CPRPC was not a useful scale in our cohort. In contrast, we used a more accurate score to predict the occurrence of PPCs in our cohort. It is based on age, smoking status and predicted postoperative forced expiratory volume in 1 s. We propose that our formula should be externally validated. |
Databáze: | OpenAIRE |
Externí odkaz: |