Constructing risk prediction model of psychological distress in lung cancer patients based on structural equation model

Abstract:

Objective：To develop the risk prediction model of psychological distress in lung cancer patients by using structural equa－tion model. Methods：A total of 441 hospitalized patients with lung cancer were randomly selected and questionnaires or scales were used to investigate the general information，psychological distress，stigma，illness perception，social support，perceived stress，coping style，mindfulness and self-esteem. Structural equation model was used to analyze the association between those factors and psycho－logical distress. Results：The structural equation model for psychological distress in lung cancer patients had a good fitting degree and its results showed that social support had a direct negative effect（β=-0.540，P=0.001） and an indirect negative effect（β=-0.003） on psychological distress，perceived stress had a direct negative effect（β=0.181，P=0.001） and an indirect negative effect（β=0.004） on psychological distress，coping style had a direct positive effect（β=0.114，P=0.005），mindfulness（β=-0.021） and stigma（β=-0.201） had indirect negative effect on psychological distress，educational level had a direct（β=0.147，P=0.003） and an indirect positive effect（β=0.026） on psychological distress，family income had a direct negative effect（β=-0.105，P=0.027） and an indirect positive effect（β=0.108） on psychological distress，age had a direct negative effect（β=-0.101，P=0.019） on psychological distress，occupational status had a positive effect（β=0.094，P=0.032） on psychological distress，and drinking history（β=0.031） and cancer stage（β=0.031） had an indirect effect on psychological distress. Conclusion：The risk prediction model which was con－structed in the present study is acceptable for patients. Path analysis of all risk factors suggests that the psychological dis－tress of lung cancer patients is influenced by many different factors in different ways and he social support，coping style，mindfulness，stigma，age，family income，educational level，occupational status，drinking history and cancer stage are key predictors