Chinese Journal of Lung Cancer (Sep 2014)

Value of Modified Possum Scoring System on Predicting Operation Risk 
in Elderly NSCLC Patients

  • Rong WANG,
  • Dewei GAO,
  • Weiqin GONG,
  • Zhiru LIANG

DOI
https://doi.org/10.3779/j.issn.1009-3419.2014.09.05
Journal volume & issue
Vol. 17, no. 9
pp. 669 – 673

Abstract

Read online

Background and objective For the assessment of elderly patients can tolerate lung cancer operation, there is no clear standard. To evaluate the clinical validity of POSSUM (Physiological and Operative Severity Score for the Umeration of Mortality and Morbidity) in elderly non-small cell lung cancer (NSCLC) surgery patients, we want to provide an important basis for operation treatment decisions. Methods A total of 138 patients, with 88 males and 50 females, with elderly NSCLC surgery between December 2007 and December 2013, are included in PLA general hospital. Using the multivariate Logistic regression analysis, we evaluate the value of each factor on the actual postoperative complications mortality and morbidity. The scorings on standard POSSUM and modified POSSUM in the complication group are compared with the non-complication group using the group t test. Drawing receiver operating characteristic (ROC) curve in standard POSSUM group and modified POSSUM group, calculating the area under the curve (AUC), AUC in standard group is compared with modified group using t test. Judge if the modified POSSUM prediction is consistent with the actual mortality and morbidity. Results Among 138 patients, there were 77 postoperative complications in 59 patients, 2 cases of death. According to the Logistic regression analysis, 17 of 18 factors in standard POSSUM, pulmonary function, different TNM stage are predictors for postoperative complications (P<0.05). Age is a predictor for postoperative death (P<0.05). In the standard POSSUM scoring, actual complication group compared with non-complication group, the difference is statistically significant (P<0.01). In the modified POSSUM scoring, complication group is compared with non-complication group, the difference is statistically significant (P<0.01). Compared with the standard POSSUM, the modified POSSUM has better predictive value on postoperative morbidity, and the comparison of AUC between the two groups is statistically significant. But the latter shows the overpredicted mortality (P<0.01). Conclusion The modified POSSUM has a good predictive value on postoperative complications in elderly NSCLC surgery patients, so it can provide the basis for decision-making operation treatment.

Keywords