de Bosset Vanessa, Atashili Julius, Miller William, Pignone Michael
Department of Epidemiology, School of Public Health, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599-7435, USA.
Cancer Epidemiol Biomarkers Prev. 2008 Apr;17(4):834-7. doi: 10.1158/1055-9965.EPI-07-2760.
Colorectal cancer (CRC) screening rates remain low in the United States. The effect of health insurance on CRC screening is not clear. We assessed the association between having health insurance and being screened for CRC and the factors that modify this association.
We used data from the 2005 Virginia Behavioral Risk Factor Surveillance System to evaluate the association of self-reported insurance coverage on self-reported CRC screening among all men and women ages > or =50 years (N = 2,887). Prevalence odds ratios (POR) were estimated using unconditional logistic regression. All covariates were assessed for potential effect measure modification and confounding. All analyses accounted for the Behavioral Risk Factor Surveillance System complex survey sampling design.
Overall, participants who reported having insurance coverage were more than twice as likely to report being screened for CRC compared with those who reported having none [crude POR, 2.16; 95% confidence interval (95% CI), 1.26-3.68]. This relationship differed between men and women (POR(males), 3.37; 95% CI, 1.63-6.96; POR(females), 1.46; 95% CI, 0.74-2.89). After adjusting for age and income, self-reported insurance coverage had a positive association with report of being screened among men (POR, 2.02; 95% CI, 0.96-4.23) but not among women (POR, 0.81; 95% CI, 0.34-1.93).
Men who reported having health insurance were more likely to report having CRC screening than those who reported not having insurance coverage. However, this effect was not observed in women. These findings, if confirmed in other study populations, indicate that improving CRC screening coverage may require not only insurance status specifications but also gender-explicit considerations.
在美国，结直肠癌（CRC）筛查率仍然很低。医疗保险对CRC筛查的影响尚不清楚。我们评估了拥有医疗保险与接受CRC筛查之间的关联以及影响这种关联的因素。
我们使用2005年弗吉尼亚行为风险因素监测系统的数据，评估年龄≥50岁的所有男性和女性（N = 2887）中自我报告的保险覆盖情况与自我报告的CRC筛查之间的关联。使用无条件逻辑回归估计患病率比值比（POR）。评估所有协变量的潜在效应测量修正和混杂情况。所有分析均考虑了行为风险因素监测系统的复杂调查抽样设计。
总体而言，报告有保险覆盖的参与者报告接受CRC筛查的可能性是报告没有保险覆盖者的两倍多[粗POR，2.16；95%置信区间（95%CI），1.26 - 3.68]。这种关系在男性和女性之间有所不同（男性POR，3.37；95%CI，1.63 - 6.96；女性POR，1.46；95%CI，0.74 - 2.89）。在调整年龄和收入后，自我报告的保险覆盖情况与男性中接受筛查的报告呈正相关（POR，2.02；95%CI，0.96 - 4.23），但在女性中则不然（POR，0.81；95%CI，0.34 - 1.93）。
报告有医疗保险的男性比报告没有保险覆盖的男性更有可能报告接受了CRC筛查。然而，在女性中未观察到这种效应。如果在其他研究人群中得到证实，这些发现表明，提高CRC筛查覆盖率可能不仅需要考虑保险状况，还需要考虑性别因素。
