Prognostic factors in gastric cancer evaluated by using Cox regression model.

Objective: To identify the most relevant short-term predictor variables in gastric cancer removal.

Experimental Design: A retrospective survival analysis executed by using the Cox regression model; the follow-up period is included between 18 and 90 months.

Setting: A district general hospital surgery unit: "Divisione di Chirurgia Generale, Ospedale Civile di Urbino" (Marche, Italy).

Subjects: One hundred and twenty nine consecutive patients operated for gastric cancer.

Interventions: Surgery (total or subtotal gastrectomy).

Main Outcome Measure: Survival times.

Results: Lymph node involvement (N) (p < 0.0001), cancer intraparietal extension (T) (p < 0.001) and the age of the patients (p < 0.05) have been recognized as significant prognostic factors.

Conclusions: Results show that the short-term prognosis largely depends on both the earliness of the diagnosis and the age of the patients.