Periodontal risk calculator versus periodontal risk assessment

Developing and testing a prediction model for periodontal disease using machine learning and big electronic dental record data

Despite advances in periodontal disease (PD) research and periodontal treatments, 42% of the US population suffer from periodontitis. PD can be prevented if high-risk patients are identified early to provide preventive care. Prediction models can help assess risk for PD before initiation and progression; nevertheless, utilization of existing PD prediction models is seldom because of their suboptimal performance. This study aims to develop and test the PD prediction model using machine learning (ML) and electronic dental record (EDR) data that could provide large sample sizes and up-to-date information. A cohort of 27,138 dental patients and grouped PD diagnoses into: healthy control, mild PD, and severe PD was generated. The ML model (XGBoost) was trained (80% training data) and tested (20% testing data) with a total of 74 features extracted from the EDR. We used a five-fold cross-validation strategy to identify the optimal hyperparameters of the model for this one-vs.-all multi-class classification task. Our prediction model differentiated healthy patients vs. mild PD cases and mild PD vs. severe PD cases with an average area under the curve of 0.72. New associations and features compared to existing models were identified that include patient-level factors such as patient anxiety, chewing problems, speaking trouble, teeth grinding, alcohol consumption, injury to teeth, presence of removable partial dentures, self-image, recreational drugs (Heroin and Marijuana), medications affecting periodontium, and medical conditions such as osteoporosis, cancer, neurological conditions, infectious diseases, endocrine conditions, cardiovascular diseases, and gastroenterology conditions. This pilot study demonstrated promising results in predicting the risk of PD using ML and EDR data. The model may provide new information to the clinicians about the PD risks and the factors responsible for the disease progression to take preventive approaches. Further studies are warned to evaluate the prediction model's performance on the external dataset and determine its usability in clinical settings.

Periodontal risk assessment in a teaching hospital population in Saudi Arabia's Eastern Province

OBJECTIVE

With this cross-sectional study, we aimed to evaluate factors associated with moderate and high risk of periodontal disease (PD) progression in the Saudi population.

METHODS

We reviewed 281 patients' clinical charts from predoctoral periodontal clinics at the dental teaching hospital in the College of Dentistry (COD) at Imam Abdulrahman Bin Faisal University (IAU) in Dammam, Saudi Arabia. After obtaining ethical approval, we determined the Periodontal Risk Assessment (PRA) of the included patients based on the modified criteria developed by Lang and Tonetti (2003). We used logistic regression on stratified data and divided the results into two categories (low-moderate and high risk) to assess the effect modifier for potential risk factors. We used SPSS version 22 for data analysis, and considered a P-value ≤ 0.05 to be statistically significant.

RESULTS

Out of the 281 patients, 104 (37.0%) were male and 177 (63.0%) were female, with a mean age of 39.9 ± 14.0 years; 78.1% were Saudi nationals, 77% were married, and 44.6% were in the age group of 30 to 49. The PRA revealed 86 (30.5%) to represent high risk, 108 (38.3%) denoted moderate risk, and 88 (31.2%) signaled low risk for periodontitis. Logistic regression analysis showed that males were three times more likely to have high PRA (OR = 3.24) and to be married (OR = 2.77), as well as to be active smokers (OR = 8.87). The highest predictive factors of high PRA were 8 or more pockets ≥ 5 mm (OR = 29.0), those with active diabetes mellitus (DM; OR = 10.2), and those with 8 or more missing teeth (OR = 9.15).

CONCLUSION

Saudi males who are married and have residual periodontal pockets, are actively diabetic, and with missing teeth are at high risk of PD. Further research is needed with a larger sample size comparing the general population with and without PD.

Long term comparison of the prognostic performance of PerioRisk, periodontal risk assessment, periodontal risk calculator, and staging and grading systems

BACKGROUND

Clinicians predominantly use personal judgment for risk assessment. Periodontal risk assessment tools (PRATs) provide an effective and logical system to stratify patients based on their individual treatment needs. This retrospective longitudinal study aimed to validate the association of different risk categories of four PRATs (Staging and grading; Periodontal Risk Assessment (PRA); Periodontal Risk Calculator; and PerioRisk) with periodontal related tooth loss (TLP), and to compare their prognostic performance.

METHODS

Data on medical history, smoking status, and clinical periodontal parameters were retrieved from patients who received surgical and non-surgical periodontal treatment. A comparison of the rate of TLP and non-periodontal related tooth loss (TLO) within the risk tool classes were performed by means of Kruskal-Wallis test followed by post-hoc comparison with the Bonferroni test. Both univariate and multivariate Cox Proportional hazard regression models were built to analyze the prognostic significance for each single risk assessment tool class on TLP.

RESULTS

A total of 167 patients with 4321 teeth followed up for a mean period of 26 years were assigned to four PRATs. PerioRisk class 5 had a hazard ratio of 18.43, Stage 4 had a hazard ratio of 7.99, and PRA class 3 had a hazard ratio of 6.13 compared with class/stage I. With respect to prognostic performance, PerioRisk tool demonstrated the best discrimination and model fit followed by PRA.

CONCLUSION

All PRATs displayed very good predictive capability of TLP. PerioRisk showed the best discrimination and model fit, followed by PRA.