Preliminary Study On A Two-stage Muti-cancer Risk Assessment Model For Rural Anhui

Objective This study aims at exploring a two-stage mult-cancer risk assessment model via integrating the results of systematic review, hospital-based and population-based case-control studies to identify high-risk individuals for developing cancers in rural areas.Methods Systematic review was implemented before investigation; Stata12.0 was used to calculate pooled odds ratio of each influence factor; during November 2013 to now, patients with cancer in situ, coming from rural Lu’an, were chosen as cases in five public hospitals in Hefei, Anhui and controls were chosen from rural areas of Lu’an city. Investigations were conducted face to face using network questionnaires; descriptive analysis, t test, chi square test, univariate and multi-variate analysis were conducted via SPSS17.0; investigation was under high quality control. We strictly followed the ethics regulations, and all informed consents were signed.Result The preliminary cancer assessment consisted of a rapid risk assessment(RRA) and a detailed risk assessment(DRA), with RRA containing 191 factors and DRA 936 questions. The preliminary study collected 115 eligible breast cancer cases and 104 cervical cancer cases with 345 breast cancer controls and 312 cervical cancer controls. No significant differences were observed in distributions of both age and education. The risk factors of breast cancer for rural females in Anhui include breast masses(OR 16.036; 95%CI 7.333-35.068), hyperplasia of mammary glands(OR 2.601; 95%CI 1.124-6.020), abortion(OR 3.115; 95%CI 1.538-6.309), long-term oral contraceptive use(OR 2.988; 95%CI 1.087-8.212) and long-term spicy food intake(OR 2.698;95%CI 1.535-4.740), protective factors are tubal ligation(OR 0.298; 95%CI 0.165-0.506) and long-term garlic intake(OR 0.507; 95%CI 0.297-0.866); the risk factors of cervical cancer contains abnormal vaginal bleeding(OR 82.129; 95%CI 19.266-350.100), abdominal pain or straining feeling(OR 5.268; 95%CI 1.388-19.900), repeatedly abnormal leucorrhea(OR 3.304; 95%CI 1.415-7.714), cervical cyst(OR 27.451; 95%CI 1.522-495.262), abortion(OR 3.319; 95%CI 1.881-5.859) and long-term spicy food intake(OR 3.296; 95%CI 1.366-7.954), protective factor for cervical cancer is tubal ligation(OR 0.218; 95%CI 0.069-0.693). The area under the receiver operating characteristic curve(AUC) of breast cancer risk assessment model I is 0.875 while the AUC of model II is 0.799; The AUC of cervical cancer assessment model I is 0.886 and the AUC of model II is 0.763.Conclusion The risk factors of female breast cancer in Lu’an, Anhui included breast masses, hyperplasia of mammary glands and long-term oral contraceptive use, while the protective factor was long-term garlic intake. The cervical cancer risk factors contained abnormal vaginal bleeding, abdominal pain or straining feeling, repeatedly abnormal leucorrhea, cervical cyst and early menarche. Abortion and long-term spicy food intake were the common risk factors of two cancers with tubal ligation being common protective factor. This study built two types of breast/ cervical cancer risk assessment models for rural females in Lu’an, Anhui province and these models are relatively accurate.