Restricting Periodontal Treatment Frequency: Impact on Tooth Loss in Danish Adults

Abstract

Objective: The study aimed to estimate the effect of a periodontal treatment policy that would restrict the receipt of periodontal therapy to no more than once every second year, on the 10-year risk of tooth extraction among Danish adults.

Methods: Data from linked nationwide Danish registers consisted of a random sample of 20 000 50-year-olds who were followed from the beginning of 1990 to the end of 2021. The longitudinal modified treatment policies' causal inference framework was used. In each of two slightly different counterfactual scenarios, the receipt of supragingival or subgingival periodontal therapy was restricted to no more than once every second year. The cumulative incidence of tooth extraction from 2012 to 2021 was compared between the counterfactual scenarios and the observed periodontal visiting pattern, while informative censoring, and time-varying and time-invariant confounding were accounted for using the social, economic and dental service utilisation history.

Results: During the 10-year follow-up period, 5021 (25.1%) individuals received at least one tooth extraction. In the two counterfactual scenarios, the number of years receiving supragingival or subgingival periodontal therapy was 30%-50% lower than in the observed data. The 10-year cumulative incidence of tooth loss was practically the same in the two counterfactual scenarios as under the observed periodontal visiting patterns.

Conclusion: The findings indicate that a considerable decrease in the number and frequency of periodontal care visits would not have significant impact on the incidence of tooth loss in Denmark.

Keywords: Dental services research; Epidemiology; Periodontitis; Public health policy; Statistics.

FIGURE 1

A simplified directed acyclic graph (DAG). C: baseline covariates, X: receipt of periodontal therapy, Y: tooth extraction, L: time‐varying covariates preceding the exposure each time point (income, diabetes) and W: time‐varying covariates not preceding the exposure each time point (restorations, examinations, radiographs, endodontics). Numbers refer to three timepoints (1–3). Please note that for the sake of clarity, not all assumed causal relationships are shown. It was thus assumed (but not shown) that L and W were also affecting Y, X, L and W with over‐year lags. In addition, C was assumed to affect all other nodes. Dashed lines indicate the causal paths of interest.

FIGURE 2

The proportions of people who received subgingival or surgical periodontal treatments or only supragingival treatment as observed between 2011 and 2020, and under the two counterfactual scenarios with modified periodontal treatment policy. Scenario 1: Individuals could not receive any periodontal care in two consecutive years. Scenario 2: Individuals could not receive supragingival care in two consecutive years but could receive subgingival or surgical periodontal care as actually observed. The curves for Scenario 2 and Observed are therefore identical for the receipt of subgingival or surgical periodontal care.

FIGURE 3

Differences in the estimated cumulative incidence of at least one non‐surgical tooth extraction between two modified treatment policies (Scenarios 1 and 2) and the observed periodontal treatment patterns in terms of cumulative incidence curves (A), incidence difference (B) and incidence ratio (C). The shaded areas represent 95% confidence intervals. Scenario 1: Individuals could not receive any periodontal care in two consecutive years. Scenario 2: Individuals could not receive supragingival care in two consecutive years but could receive subgingival or surgical periodontal care as actually observed.

FIGURE 4

The proportions of people who received subgingival or surgical periodontal treatments or only supragingival treatment as observed between 2011 and 2020, and under the counterfactual scenario with modified periodontal treatment policy. Scenario 3: Individuals were more likely to receive periodontal care in consecutive years.

FIGURE 5

Differences in the estimated cumulative incidence of at least one non‐surgical tooth extraction between modified treatment policy (Scenario 3) and the observed periodontal treatment patterns in terms of cumulative incidence curves (A), incidence difference (B) and incidence ratio (C). The shaded areas represent 95% confidence intervals. Scenario 3: Individuals were more likely to receive periodontal care in consecutive years.