[Analysis of influencing factors on orthodontic efficacy in adolescent patients and construction of logistic regression model].

PURPOSE: To analyze the influencing factors of orthodontic efficacy in adolescent patients and construct a logistic regression model.

METHODS

A total of 260 adolescent orthodontic patients treated from January 2019 to December 2023 were selected. The therapeutic effect was evaluated using the peer assessment rating (PAR) index. At the 2-year follow-up after treatment completion, patients with a reduction of more than 5 points in the PAR index compared with immediately post-treatment were assigned to the poor efficacy group (n=53), and the remaining patients were assigned to the favorable efficacy group (n=207). Binary logistic regression was used to analyze the risk factors for poor efficacy in adolescent orthodontic patients, and logistic regression model was constructed based on the risk factors.

RESULTS

The binary logistic regression analysis showed that anxiety (OR=8.467, 95%CI=2.800-25.609, P＜0.001), Angle Class Ⅱ malocclusion (OR=9.624, 95%CI=2.504-36.990, P=0.001), Angle Class Ⅲ malocclusion (OR=4.210, 95%CI=1.146-15.467, P=0.030), root morphology abnormalities (OR=5.927, 95%CI=2.276-15.435, P＜0.001), poor oral hygiene (OR=2.943, 95%CI=1.139-7.604, P=0.026) and poor patient compliance (OR=9.620, 95%CI =3.047-30.376, P＜0.001) were risk factors for poor treatment outcome in adolescent orthodontic patients, while retainer wearing time (OR=0.773, 95%CI=0.69-0.864, P＜0.001) was a protective factor. A logistic regression model was constructed using the risk factors. The Hosmer-Lemeshow goodness-of-fit test showed that P=0.575. The area under the curve (AUC) of the model for predicting poor treatment outcomes in adolescent orthodontic patients was 0.939, with a 95%CI of 0.908-0.971. The sensitivity and specificity were 92.5% and 83.6%, respectively, and the practical application accuracy was 89.2%.

CONCLUSIONS

The orthodontic efficacy of adolescent orthodontic patients is related to anxiety, types of malocclusions, abnormal root morphology, oral hygiene, compliance and duration of retainer wearing. A logistic regression model constructed through risk factors has high predictive value.