Objective  To evaluate the efficacy of radiotherapy, surgery and combined therapy among gastric cancer patients with inverse probability of treatment weighting (IPTW) method.

  Methods  We extracted the data on 7 005 pathogenically diagnosed gastric cancer patients (3 983 undergoing surgery, 795 with radiotherapy, and 2 227 receiving combined therapy) from the datasets of Surveillance, Epidemiology, and End Results Program (SEER) of United States for the period from January 2004 through 2013. Then we conducted a retrospective analysis on survival and therapeutic outcome of the patients with IPTW method.

  Results  For the patients with surgery, radiotherapy, and the combined therapy, the median survival time were 30, 9, and 38 months before weighted with IPTW and 25, 11, and 38 months after weighted with IPTW. There were significant differences in survival curves for the three groups of patients before and after weighted with IPTW (P < 0.001 for all). Patients with the combined therapy were better for prognosis than those with surgery or radiotherapy before and after weighting analysis with IPTW. The absolute standardized mean difference (ASMD) for survival analysis results weighted by IPTW for the three patients groups were all less than 0.2, indicating a balanced weighting effect among the three patient groups. Cox multivariate regression analysis demonstrated that compared with those with surgery, the patients with radiotherapy had a poor prognosis (hazard risk HR= 2.044, 95% confidence interval 95% CI: 1.770 – 2.361), whereas the patients with the combined therapy had a better prognosis (HR = 0.630, 95% CI: 0.573 – 0.694) after adjusting for confounders such as gender, age, race, marital status, tumor size, pathogenically differentiated degree, tumor stage, and tumor-node-metastasis (TNM) staging.

  Conclusion   T he prognosis of gastric cancer patients with combined therapy is better than the patients with surgery or radiotherapy.