Background:
A 10-year survival analysis was performed to assess the predictive validity of the periodontal risk score (formerly known as Miller-McEntire Periodontal Prognostic Index [MMPPI]) to predict long-term survival of periodontally diseased molars in a longitudinally assessed cohort.
Materials and Methods:
The MMPPI scores were computed at baseline by summarizing scores allocated for individual factors. The cohort was treated, followed up, and placed under regular periodontal maintenance. Molar survival data collected up to 10 years of follow-up were analyzed. Cox proportional regression was performed, and hazards ratio (HR) were computed for each prognostic factor and the MMPPI score. To estimate the predictive value of MMPPI, a receiver operating curve (ROC) curve analysis was performed.
Results:
From 1032 molars, 155 molars were extracted over the 10-year follow-up duration. Cox proportional hazard analysis showed significant hazard ratios for tooth loss for the component variables significant HR was noted for age: 4.92 (3.34:7.27), smoking: 1.74 (1.38–2.22), diabetes: 1.66 (1.49–1.86), molar type: 1.39 (1.15–1.67), probing depth: 2.00 (1.63–2.46), furcation: 2.64 (2.30–3.03), mobility: 3.45 (2.98–4.01), and total MMPPI score: 1.98 (1.85–2.12). ROC curve analysis showed an area under the curve value of 0.94 for the MMPPI index as a predictor of molar loss at 8 years, and the Youden index was maximized at the optimal cutoff point score of 7.
Conclusions:
All component scores of MMPPI showed significant hazard ratios at 10 years. These findings support the previous results from the 5-year analysis of this university-based cohort and warrant validation in independent cohorts.