OBJECTIVES: The Nationwide Inpatient Sample is part of the Healthcare Cost and Utilization Project, and is the only national hospital database with charge information on all patients, regardless of payer, including persons covered by Medicare, Medicaid, private insurance, and the uninsured. It is the purpose of this study to examine the relationship between patient outcomes and conditions of the patients undergoing different treatments for lung cancer. METHODS: There are fifteen possible patient diagnoses in the dataset. SAS Enterprise Guide was used to obtain Lung Cancer data from NIS by using the CATX and the RXMATCH statements in SAS. We bring all fifteen diagnoses into one column as a string of codes, using the CATX function. Total charges are used to examine the relationship between diagnoses and procedures. The generalized linear regression model was used to fit the data. RESULTS: After filtering down to lung cancer using the strings of diagnoses, there were 5457 records in the data set. By the plot method, we selected variables related to Total charges. We found that the Total charges were highly related to Age in years at admission, Diagnosis Related Group, Length of stay and Died during hospitalization. By the basic criterion, deviance residuals and Pearson chi-square residuals, the specified model fits the data reasonably well. From the Type 1 and Type 3 analysis, all the estimates for the intercept, Los, Age, DRG, and Died were 10.1805, 0.1181, 0.003, -0.0008, and -1.024, respectively. All of them were statistically significant. Co-morbid diagnoses that increased total charges include coronaries, multiple significant traumas, and cardiac implant. CONCLUSIONS: Our analysis revealed that there was a specified relationship between these variables. With increasing of length of stay, age in years at admission, small codes of diagnosis related group and surviving during hospitalization, the total charges increase, which is reasonable.