Meta-analysis Of Case-control Studies Of Diabetes And Pancreatic Cancer

Objective Using the method of Meta（meta-analysis）evaluate theassociation between diabetes and pancreatic cancer, and provide the basis forthe early diagnosis and prevention of pancreatic cancer.Methods Using original literatures, and conducting searches viadatabases of PubMed、EMbase, CNKI and CBMdisc, etc., all materialswere dated up July of2013，At the same time, to retrieve relevantliterature references. All case-control studies that associate with diabetes andpancreatic cancer were included and the quality of them was assessed. Afterscreening, the quality evaluation of literatures， Meta-analysis wasconducted using Review Manager5.0software.Results A total of21case-control studies were included, Screeningby Jadad scoring system,18met the inclusion criteria for Meta-analysis,including6184cases of patients with pancreatic cancer group and10296cases of control group. As compared with the control group, the results ofpancreatic cancer group by Meta-analysis showed that the result ofheterogeneity test from all research data is I2=74%，χ2=65.38，P<0.05, therandom effects model will be adopted, the pooled odds ratio (OR) was3.60 and95%confidence interval (95%CI) was3.28-3.94, there will be somedifference in statistics，pancreatic cancer patients’ diabetes risk was3.6times than the normal population. Conclusion Diabetes was associatedwith pancreatic cancer，diabetes may be an early clinical manifestation ofpancreatic cancer.