LAUSR

INTRODUCTION Tooth extraction under general anesthetic (GA) is a global health problem. It is expensive, high risk, and resource intensive, and its prevalence and burden should be reduced where possible. Recent innovation in data analysis techniques now makes it possible to assess the impact of GA policy decisions on public health outcomes. This article describes results from one such technique called process mining, which was applied to dental electronic health record (EHR) data. Treatment pathways preceding extractions under general anesthetic were mined to yield useful insights into waiting times, number of dental visits, treatments, and prescribing behaviors associated with this undesirable outcome. METHOD Anonymized data were extracted from a dental EHR covering a population of 231,760 patients aged 0 to 16 y, treated in the Irish public health care system between 2000 and 2014. The data were profiled, assessed for quality, and preprocessed in preparation for analysis. Existing process mining methods were adapted to execute process mining in the context of assessing dental EHR data. RESULTS Process models of dental treatment preceding extractions under general anesthetic were generated from the EHR data using process mining tools. A total of 5,563 patients who had 26,115 GA were identified. Of these, 9% received a tooth dressing before extraction with an average lag time of 6 mo between dressing and extraction. In total, 11,867 emergency appointments were attended by the cohort with 2,668 X-rays, 4,370 prescriptions, and over 800 restorations and other treatments carried out prior to tooth extraction. DISCUSSION AND CONCLUSIONS Process models generated useful insights, identifying metrics and issues around extractions under general anesthetic and revealing the complexity of dental treatment pathways. The pathways showed high levels of emergency appointments, prescriptions, and additional tooth restorations ultimately unsuccessful in preventing extractions. Supporting earlier publications, the study suggested earlier screening, preventive initiatives, guideline development, and alternative treatments deserve consideration. KNOWLEDGE TRANSFER STATEMENT This study generates insights into tooth extractions under general anesthetic using process mining technologies and methods, revealing levels of extraction and associated high levels of prescriptions, emergency appointments, and restorative treatments. These insights can inform dental planners assessing policy decisions for tooth extractions under general anesthetic. The methods used can be combined with costs and patient outcomes to contribute to more effective decision-making.