O Rhouma1, A D McMahon, R R Welbury. 1. Department of Paediatric Dentistry, Glasgow Dental Hospital and School, Glasgow, UK. ousamarhouma@hotmail.com
Abstract
AIM: To identify early clinical variables that are most predictive of treatment outcome and to develop a model that will allow prediction of treatment outcomes based on these variables. STUDY DESIGN AND METHODS: A dental trauma database was used to randomly identify patients who had received treatment for avulsed teeth between 1998 and 2007. A data extraction form was designed and completed for each tooth. Demographic, diagnostic and treatment information recorded in the patient's records, in addition to radiographs, were viewed retrospectively. STATISTICS: The significance and the predictive power for each early clinical variable were assessed using a univariate logistic regression model. Only significant variables (p<0.05) were considered eligible for the prediction model and a c-index was then constructed for their respective predictive power (0.5 = no predictive power, 1.0 = perfect prediction). RESULTS: Of the original sample of 213 patients who had received treatment for avulsed teeth between 1998-2007 only 105 fulfilled the criteria for evaluation. Two models ('At first visit' and 'at initial treatment visits') were produced with a total of five variables that were significant and holding the greatest predictive power (high c-index): patient age (p=0.001, c=0.80); stage of root formation (p=0.001, c=0.76); storage medium (p=0.047, c=0.58); tooth mobility after dressing (p=0.001, c=0.70); and tooth mobility after splinting (p=0.003, c=0.70). These variables underwent multi-variate analysis and the final models had good predictive abilities (c-index of 0.80 and 0.74). CONCLUSION: These predictive models based on patient age, stage of root formation, storage medium, tooth mobility after dressing and tooth mobility after splinting were shown to have high predictive value and will enable a clinician to estimate the long term prognosis of avulsed and replanted teeth. It will enable planning for further treatment with a realistic view of outcome at an early stage.
AIM: To identify early clinical variables that are most predictive of treatment outcome and to develop a model that will allow prediction of treatment outcomes based on these variables. STUDY DESIGN AND METHODS: A dental trauma database was used to randomly identify patients who had received treatment for avulsed teeth between 1998 and 2007. A data extraction form was designed and completed for each tooth. Demographic, diagnostic and treatment information recorded in the patient's records, in addition to radiographs, were viewed retrospectively. STATISTICS: The significance and the predictive power for each early clinical variable were assessed using a univariate logistic regression model. Only significant variables (p<0.05) were considered eligible for the prediction model and a c-index was then constructed for their respective predictive power (0.5 = no predictive power, 1.0 = perfect prediction). RESULTS: Of the original sample of 213 patients who had received treatment for avulsed teeth between 1998-2007 only 105 fulfilled the criteria for evaluation. Two models ('At first visit' and 'at initial treatment visits') were produced with a total of five variables that were significant and holding the greatest predictive power (high c-index): patient age (p=0.001, c=0.80); stage of root formation (p=0.001, c=0.76); storage medium (p=0.047, c=0.58); tooth mobility after dressing (p=0.001, c=0.70); and tooth mobility after splinting (p=0.003, c=0.70). These variables underwent multi-variate analysis and the final models had good predictive abilities (c-index of 0.80 and 0.74). CONCLUSION: These predictive models based on patient age, stage of root formation, storage medium, tooth mobility after dressing and tooth mobility after splinting were shown to have high predictive value and will enable a clinician to estimate the long term prognosis of avulsed and replanted teeth. It will enable planning for further treatment with a realistic view of outcome at an early stage.