Abstract: Objective To understand the status and risk factors of housebound elderly people in community and to construct a risk prediction model for them. Methods A total of 587 housebound elderly people in 3 communities in Qingdao were investigated by using general information questionnaire,stay at home,Barthel Index, Geriatric Depression Scale (GDS), and Social Support Rating Scale (SSRS). Univariate analysis and logistic regression were used to establish a risk prediction model. The area under the ROC curve was used to test the model. Results Of the 587 objects, 107 of them were housebound elderly people, accounting for 18.2%.The study finally included elder age (age 81~96,OR=3.062),low income（OR=33.199）,lack of activity hobbies（OR=4.285）,lack of physical exercise(sometimes,OR=9.457;no exercise,OR=32.432）,severe dysfunction（OR=7.609）, and depression（OR=16.050）to construct the risk prediction model. The area under the ROC curve of this model was 0.937,with the sensitivity of 0.944,the specificity of 0.765 and the Youden of 0.709. Conclusion The prevalence of housebound is relatively high, and there are many risk factors. The risk prediction model established has a satisfactory prediction effect, which can be used as a risk screening tool for housebound elderly people in community.

Key words: community, elderly people, housebound, risk prediction model