In Haejin, Langdon-Embry Marisa, Gordon Lauren, Schechter Clyde B, Wylie-Rosett Judith, Castle Philip E, Margaret Kemeny M, Rapkin Bruce D
Montefiore Medical Center/Albert Einstein College of Medicine, Department of Surgery, Bronx, New York; Albert Einstein College of Medicine, Department of Epidemiology and Population Health, Bronx, New York.
Montefiore Medical Center/Albert Einstein College of Medicine, Department of Surgery, Bronx, New York.
J Surg Res. 2018 Jul;227:246-256. doi: 10.1016/j.jss.2018.02.053. Epub 2018 Apr 3.
A questionnaire that distinguishes how variability in gastric cancer prevalence is associated with ethnicity/birth country/immigration/cultural diet along with known risk factors may improve targeting populations for gastric cancer screening in the United States.
Existing literature was used to identify the item pool. Cluster analysis, focus groups, and cognitive interviewing were used to reduce collinear items and refine the questionnaire. Logistic regression analysis was used to determine which items distinguished gastric cancer cases from the primary care and community controls.
The results of analysis of data from 40 cases and 100 controls (primary care = 47; community = 53) were used to reduce the 227 item pool to 12 items. After ranking these variables using model bootstrapping, a logistic regression model using the highest ranked eight variables was chosen as the final model. Older age, foreign nativity, daily consumption of cultural food at ages 15-18, less than high-school education, and greater acculturation were significantly associated with being a gastric cancer case compared with the controls.
An eight-item survey that addresses gastric cancer risk factors, ethnicity, cultural habits, and immigration patterns has potential to identify high-risk persons from multicultural areas within the US, who might benefit from endoscopic screening for gastric cancer.
一份能够区分胃癌患病率的变异性如何与种族/出生国家/移民/文化饮食以及已知风险因素相关联的问卷，可能会改善美国胃癌筛查目标人群的确定。
利用现有文献确定项目池。采用聚类分析、焦点小组和认知访谈来减少共线项目并完善问卷。使用逻辑回归分析来确定哪些项目能够区分胃癌病例与初级保健和社区对照人群。
对40例病例和100例对照（初级保健 = 47例；社区 = 53例）的数据进行分析的结果，用于将227项的项目池减少到12项。在使用模型自抽样对这些变量进行排序后，选择使用排名最高的8个变量的逻辑回归模型作为最终模型。与对照相比，年龄较大、外国出生、15 - 18岁时每日食用文化食物、高中以下学历以及更高的文化适应程度与胃癌病例显著相关。
一份涉及胃癌风险因素、种族、文化习惯和移民模式的八项调查问卷，有潜力识别美国多元文化地区的高危人群，这些人可能受益于胃癌内镜筛查。
