Managing Dental Appointment No-Shows: A Systematic Review of Machine Learning Applications

Missed appointments, commonly referred to as no-shows, are a persistent operational challenge in dental clinics, resulting in wasted clinical resources, disturbed workflows, and substantial financial losses. Machine learning, a predictive approach that identifies complex data patterns, has emerged as a promising tool for predicting patients at risk of missing scheduled dental visits. This systematic review aimed to synthesize and critically appraise machine learning-based prediction models developed to predict dental appointment no-shows.

A comprehensive search of PubMed, Scopus, ScienceDirect, Google Scholar, and other databases was conducted, and eligible studies were screened using predefined inclusion and exclusion criteria. Data extraction and quality assessment were performed using the prediction model risk of bias assessment tool (PROBAST).

The findings highlighted the applicability of machine learning models for no-show prediction with respect to dental appointments. However, direct comparisons across consistent datasets and evaluation metrics were limited. Key predictors included long lead times between the booking and appointment dates, missed confirmation messages, past no-show history, and temporal factors such as day of the week. Despite the reported performance of the models, existing studies were limited by small sample sizes, single-center designs, and a lack of external validation.

Overall, machine learning represents a valuable strategy for improving appointment management in dental clinics, but larger multicenter research and real-world implementation studies are required.