Patient history as a predictor of future treatment need? Considerations from a dataset containing over nine million courses of treatment

Abstract

Aim It is the aim of this paper to consider whether overall patient treatment history per se and what length of patient history, matters in predicting future treatment need. Methods This study used a data set (SN7024, available from UKDataService), consisting of treatment records for General Dental Services' (GDS) patients, this being obtained from all items of service payment records for patients treated in the GDS of England and Wales between 1990 and 2006. For the purpose of this study, the GDS dataset for patients attending in 2003 was restricted to adult patients (aged 18 or over on 31 December 1990; that is, year of birth earlier than 1973) who attended in both two-year periods 1991/2 and 2004/5. Each course of treatment was classified as 'active' (eg restoration, extraction, prosthesis) or 'not active' (eg prevention, diagnosis). Treatment costs for 2001-2005 (outcome), 2000 (one-year history), 1999-2000 (two-year history) and so on until 1991-2000 (ten-year history) were determined, and history and outcome correlated. Results A total of 455,844 patients met the inclusion criteria, namely adults with a full history. They received 9,341,583 courses of treatment, of which 49% were classified as 'active' and 51% as 'not active'. The analysis indicated that both total costs and active treatment costs are positively correlated with their historical values, with the correlation coefficients increasing from 0.24 and 0.25 with one year of history to 0.42 and 0.44 with ten years of history. Overall, therefore, future treatment cost is correlated with past treatment costs. Conclusions Treatment history may provide an important correlate of future dental treatment needs and the more history the better, at least up to five years. However, active treatment is the important component and should be distinguished from preventive and diagnostic treatments.