Predicting periodontitis progression?

Abstract

Cochrane Library, Ovid, Medline, Embase and LILACS were searched using no language restrictions and included information up to July 2014. Bibliographic references of included articles and related review articles were hand searched. On-line hand searching of recent issues of key periodontal journals was performed (Journal of Clinical Periodontology, Journal of Dental Research, Journal of Periodontal Research, Journal of Periodontology, Oral Health and Preventive Dentistry). Prospective and retrospective cohort studies were used for answering the question of prediction since there were no randomised controlled trials on this topic. Risk of bias was assessed using the validated Newcastle-Ottawa quality assessment scale for non-randomised studies. Cross-sectional studies were included in the summary of currently reported risk assessment tools but not for risk of progression of disease, due to the inability to properly assess bias in these types of studies. Titles and abstracts were scanned by two reviewers independently.Full reports were obtained for those articles meeting inclusion criteria or those with insufficient information in the title to make a decision. Any published risk assessment tool was considered. The tool was defined to include any composite measure of patient-level risk directed towards determining the probability for further disease progression in adults with periodontitis. Periodontitis was defined to include both chronic and aggressive forms in the adult population. Outcomes included changes in attachment levels and/or deepening of periodontal pockets in millimeters in study populations undergoing supportive periodontal therapy. Data extraction was performed independently and in collaboration by two reviewers; completed evidence tables were reviewed by three reviewers. Studies were each given a descriptive summary to assess the quantity of data as well as further assessment of study variations within study characteristics. This also allowed for determining the suitability of data for further quantitative analysis (meta-analysis). Unfortunately, the heterogeneity of the data did not allow. After screening, 19 studies fitted the inclusion criteria of identifying five different patient-based periodontal risk assessment tools. DenPlan Excel/Previsor Patient Assessment (DEP-PA) and its modifications were used in five studies. The HIDEP model, the dentition risk system (DRS) and the risk assessment-based individualised treatment (RABIT) were each used in one study. Lastly, the periodontal risk assessment (PRA) and its modifications were found in 12 publications.PRA uses the following factors to assess risk of recurrence of disease: Percentage of bleeding on probing, loss of teeth from a total of 28 teeth, loss of periodontal support in relation to the patient's age, prevalence of residual pockets greater than 4 mm (3-5 mm), systemic and genetic conditions and environmental factors, such as cigarette smoking.Ten included studies had cohort designs (N= 2130) spanning three to 12 years with different follow-up times. Generally, these studies reflected that different assessment tools were able to separate subjects with differing probability of disease progression and tooth loss. The observed effect was dose dependent (the higher the estimation of risk the higher the level of observed disease or tooth loss).Six cross sectional studies (N=1078) reported the comparison of different assessment tools, adjusted or unadjusted associations with periodontal disease and subjective risk assessments provided by the tools. There were three articles noted in the flow diagram as articles proposing the tool. Qualitative analysis reflects that parameters are similar across the studies but differences are present in how these parameters were assessed. In treated populations, results of patient-based risk assessments predicted periodontitis progression and tooth loss in various populations. Additional research on the utility of risk assessment and results in improving patient management are needed.