Relationship Among Stress, Perception of Occurrence Risk of Cancer and Cancer Preventive Health Behavior in Three-Shift
This study is designed to be used as basic data for enhancing preventive health behavior against cancer of nurses by understanding stresses, perceptions of occurrence risk of and degrees of preventive health behavior against cancer of nurses working in three shifts and verifying the relationship among them.

The data of this study are collected from September 20 to October 3 in 2014. A questionnaire of only subjects who provided prior consents after a research assistant explained the purpose and method of this study, the method to protect confidentiality and fill out written consents to 252 three shifts nurses working at three medium and small hospitals with from100 to 300 beds.

For the study tools, this study uses the shortened socio-psychological stress tool developed by Jang Sejin (2000) for stress; the perception measurement tool for possibility of 5 big cancers by Yang Yeonghui (2012) for perception of occurrence risk of cancer; the preventive health behavior measurement tool against cancer derived from the other tool developed by Seo Sun-nim(1998) and corrected and complemented by Kim Ikji and Kim Sanghui (2012).

After the collected data are coded, SPSS-win 21.0 Program, t-test, ANOVA, Scheffe''s test and Pearson''s correlation analysis are used to analyze frequency, percent, average and standard deviation.

The findings are as follows:
1. Each average of subjects'' stresses, perceptions of occurrence risk of cancer and preventive health behavior against it are 24.35±7.93(0-54), 12.63±4.12(5-25) and 58.07±8.87 (18-90), respectively.

2. The stresses specific to subjects show some significant differences in age(F=7.602, p<.001), matrimonial status (t=3.732, p<.001), a total of career (F=8.954, p<.001), working career in three shifts (F=3.163, p=.044), cancer insurance status (t=3.887, p=.022), health state (F=23.220, p<.001) drinking status(t=2.706, p=.007), exercise status (t=-2.865, p=.005) and sleep disorder'' (t=7.559, p<.001) ; the perception of occurrence risk of cancer specific to them show some statistically significant differences in health state’(F=14.415, p<.001), ‘disease status’(t=3.664, p=.001), ‘family history related with cancer’(t=6.841, p=.001), ‘drinking status’(t=2.630, p=.009) and ‘sleep disorder’(t=4.345, p<.001); preventive health behavior against cancer specific to them exhibit some statistically significant differences in age (F=4.530, p=.004), ward (F=2.926, p=.034), a total of career(F=4.829, p=.009), health state(F=4.968, p=.002), drinking status(t=-6.041, p<.001), exercise status(t=3.625, p<.001) and ‘sleep disorder’(t=-4.433, p<.001).

3. The stress has a positive relationship with the perception of occurrence risk of cancer (r=.403, p<.001); the perception of occurrence risk of cancer has a positive relationship with the preventive health behavior against it (r=.313, p<.001); and the stress has a positive relationship with preventive health behavior against it(r=.313, p<.001). And their relationship is statistically significant.

The findings show that three shifts nurses'' stresses are high and degrees of perceptions of occurrence risk of and preventive health behaviors against cancer are somewhat low. And it is found that there are significant correlations among these variables. On the basis of these findings, the development of nurse intervention is required to reduce stress and enhance the other two variables in three shits nurses.