Skip to main content
. 2022 Oct 14;10:969922. doi: 10.3389/fpubh.2022.969922

Table 3.

Multivariate logistic model for poor emotional state (presence of 3–4 negative items).

Betha OR* CI 95% p-value
Type of information given to the child Realistic information 0 1
Information misrepresenting the negative aspects 0.05372 1.055 (0.712–1.565) 0.789
Creative explanations −1.48621 0.226 (0.073–0.700) 0.009
No information −0.58049 0.560 (0.227–1.378) 0.206
Appetite Has more 0 1
No change −0.30185 0.739 (0.510–1.072) 0.111
Has less 0.83570 2.306 (1.454–3.658) 0.001
Home with garden No answer 0 1
Yes −0.54739 0.578 (0.335–1.000) 0.049
No −0.10606 0.899 (0.525–1.542) 0.699
Sleep disturbance No −0.88301 0.414 (0.300–0.570) <0.001
Parent believes situation will not normalize No 1.00382 2.729 (1.791–4.157) <0.001
Only child Yes −0.40418 0.668 (0.453–0.983) 0.040
Pets in the home No 0.36440 1.440 (1.047–1.979) 0.024
Children ask about what is happening No −1.06831 0.344 (0.243–0.485) <0.001
Return to activity No −0.43769 0.646 (0.464–0.898) 0.009
Age 0.04036 1.041 (0.986–1.100) 0.148
Sex (Female) 0.16531 1.180 (0.867–1.605) 0.292

Training sample n = 1,000; test sample n = 501; number of children with a poor emotional state (3–4 items) in the training sample = 347; area under the ROC curve in the test sample = 0.812 (95% CI 0.773–0.850). Homer-Lemeshow p-value = 0.713; accuracy rate in the test sample = 75.4%. Nagelkerke's R2 = 0.339. Likelihood ratio chi-squared test = 281.8 (p < 0.001).

*

Model adjusted for control of disease.