1.
Clinical features and prognostic factors in patients with bone metastases from non-small cell lung cancer.
Wu, XT, Zhou, JW, Pan, LC, Ge, T
The Journal of international medical research. 2020;(5):300060520925644
Abstract
OBJECTIVE To investigate the clinical features and evaluate the prognostic factors in patients with bone metastases from non-small cell lung cancer (NSCLC). METHODS We retrospectively investigated 356 patients with NSCLC with bone metastases from January 2012 to December 2017. The overall survival (OS) and 1-year survival rate were calculated by Kaplan-Meier analysis and compared by univariate analysis using the log-rank test. Multivariate analysis was performed using the Cox proportional hazards model. RESULTS A total of 694 sites of bone metastases were determined among the 356 patients. The most common site of bone metastases was the ribs. The median OS was 12.5 months and the 1-year survival was 50.8% in the overall population. Univariate analysis revealed that histological type, number of bone metastases, Eastern Cooperative Oncology Group performance status (ECOG PS), bisphosphonate therapy, and serum calcium, lactate dehydrogenase, and alkaline phosphatase were significantly correlated with prognosis. Multivariate analysis identified multiple bone metastases, ECOG PS ≥2, lactate dehydrogenase ≥225 U/L, and alkaline phosphatase ≥140 U/L as independent negative prognostic factors. CONCLUSION Multiple bone metastases, high ECOG PS, and high serum alkaline phosphatase and lactate dehydrogenase are independent negative prognostic factors for bone metastases from NSCLC.
2.
Propensity score and proximity matching using random forest.
Zhao, P, Su, X, Ge, T, Fan, J
Contemporary clinical trials. 2016;:85-92
-
-
Free full text
-
Abstract
In order to derive unbiased inference from observational data, matching methods are often applied to produce balanced treatment and control groups in terms of all background variables. Propensity score has been a key component in this research area. However, propensity score based matching methods in the literature have several limitations, such as model mis-specifications, categorical variables with more than two levels, difficulties in handling missing data, and nonlinear relationships. Random forest, averaging outcomes from many decision trees, is nonparametric in nature, straightforward to use, and capable of solving these issues. More importantly, the precision afforded by random forest (Caruana et al., 2008) may provide us with a more accurate and less model dependent estimate of the propensity score. In addition, the proximity matrix, a by-product of the random forest, may naturally serve as a distance measure between observations that can be used in matching. The proposed random forest based matching methods are applied to data from the National Health and Nutrition Examination Survey (NHANES). Our results show that the proposed methods can produce well balanced treatment and control groups. An illustration is also provided that the methods can effectively deal with missing data in covariates.