Identification of children at risk of missing telemedicineappointments: development of a predictive modelduring the COVID-19 pandemic

Introduction. Although the use of telemedicine appointments has grown exponentially since the COVID-19 pandemic, missed telemedicine appointments remain a relatively understudied topic. We set out to develop and validate a predictive model to identify patients at high risk of missing telemedicine appointments. Methods. Retrospective cohort. We included telemedicine appointments from August 1, 2020, to March 31, 2021. We included as predictors the clinical characteristics of the patients, missed appointment history, appointment characteristics, social determinants of health, and weather conditions. We developed a predictive model using multivariate mixed-effects logistic regression Results. We included 3339 telemedicine appointments, with a missed appointment rate of 11.35% (95%CI 10.3-12.5). Among the risk factors for missing telemedicine appointments, we found that public health coverage (OR 2.2) and having other appointments on the same day (OR 3.2) increased the likelihood of missing telemedicine appointments. On the other hand, having a chronic condition (OR 0.5) and the number of previous appointments requested (OR 0.7) acted as protective factors. The final predictive model included 19 variables and 4 interactions, with an area under the ROC curve of 0.72 (95%CI 0.70.8) and a calibration slope of 0.78 (95%CI 0.6-0.9), indicating slight overfitting. Conclusion. In this study, we developed and validated a predictive model that identifies children at high risk of missing telemedicine appointments. This model helps guide strategies aimed at reducing missed telemedicine appointments.