Cost-effectiveness of Artificial Intelligence for Proximal Caries Detection

Abstract

Artificial intelligence (AI) can assist dentists in image assessment, for example, caries detection. The wider health and cost impact of employing AI for dental diagnostics has not yet been evaluated. We compared the cost-effectiveness of proximal caries detection on bitewing radiographs with versus without AI. U-Net, a fully convolutional neural network, had been trained, validated, and tested on 3,293, 252, and 141 bitewing radiographs, respectively, on which 4 experienced dentists had marked carious lesions (reference test). Lesions were stratified for initial lesions (E1/E2/D1, presumed noncavitated, receiving caries infiltration if detected) and advanced lesions (D2/D3, presumed cavitated, receiving restorative care if detected). A Markov model was used to simulate the consequences of true- and false-positive and true- and false-negative detections, as well as the subsequent decisions over the lifetime of patients. A German mixed-payers perspective was adopted. Our health outcome was tooth retention years. Costs were measured in 2020 euro. Monte-Carlo microsimulations and univariate and probabilistic sensitivity analyses were conducted. The incremental cost-effectiveness ratio (ICER) and the cost-effectiveness acceptability at different willingness-to-pay thresholds were quantified. AI showed an accuracy of 0.80; dentists' mean accuracy was significantly lower at 0.71 (minimum-maximum: 0.61-0.78, P < 0.05). AI was significantly more sensitive than dentists (0.75 vs. 0.36 [0.19-0.65]; P = 0.006), while its specificity was not significantly lower (0.83 vs. 0.91 [0.69-0.98]; P > 0.05). In the base-case scenario, AI was more effective (tooth retention for a mean 64 [2.5%-97.5%: 61-65] y) and less costly (298 [244-367] euro) than assessment without AI (62 [59-64] y; 322 [257-394] euro). The ICER was -13.9 euro/y (i.e., AI saved money at higher effectiveness). In the majority (>77%) of all cases, AI was less costly and more effective. Applying AI for caries detection is likely to be cost-effective, mainly as fewer lesions remain undetected. Notably, this cost-effectiveness requires dentists to manage detected early lesions nonrestoratively.

Keywords: caries diagnosis/prevention; computer simulation; decision making; dental; economic evaluation; radiology.

Figure 1.

Input data and model. The state diagram (central parts) shows the different health states (solid boxes). Transition or allocation probabilities determined the chance of passing between them, indicated by arrows. The data sources used to simulate individuals’ flow through the model are shown in dotted boxes at the left and right. Individuals started with teeth being either sound or showing E1/2/D1 and D2–D3 lesions. Sound surfaces could be detected as such, without any subsequent treatment, or false positively detected as initial (E2/D1) caries lesions depending on the detection method. False-positive detections on sound surfaces led to infiltration treatment, without any effectiveness gain, but money spent unnecessarily. Initial lesions (E2/D1) could again be detected (treated by resin infiltration) or not detected (and assumed to progress with some chance) and, depending on the efficacy of resin infiltration, be arrested or progress to D2 lesions. In a sensitivity analysis, we assumed all detected lesions to be treated restoratively instead. For advanced lesions (D2) not extending into the inner third of the dentin (D3), a two-surfaced restoration was assumed to be placed. This placement was not to be associated with pulpal risks, mainly as we assumed these lesions to be not deep (no proximity to the pulp). For lesions extending into the inner one-third of the dentin, the risk of pulp exposure was estimated at 0.3, and exposed pulps received direct pulp capping. The risk of restorative complications was derived from previous studies, and if restorations failed, they were assumed to be either renewed or repaired. If failing again, the placement of a full-metal crown (the standard crown therapy for most posterior teeth within statutory German health insurance) was assumed. Failed crowns were assumed to be replaced once, after which the tooth was extracted. Extracted teeth were assumed to be replaced using implant-supported single crowns; the proportion of replaced teeth was 0.8 in the base case and varied in sensitivity analyses.

Figure 2.

Cost-effectiveness plane, incremental cost-effectiveness, and net-benefit analysis of the base case. (A) The costs and effectiveness of the 2 comparators are plotted for 1,000 sampled individuals in each group. (B) The incremental costs and effectiveness of artificial intelligence (AI) compared with no AI are plotted. Quadrants indicate comparative cost-effectiveness (e.g., upper right: higher costs at higher effectiveness; lower right: lower costs and higher effectiveness). Inserted cross-tabulation: Percentage of samples lying in different quadrants. (C) We plotted the probability of comparators being acceptable in terms of their cost-effectiveness depending on the willingness-to-pay threshold of a payer. The range of willingness to pay was expanded from 0 to 100 euro and did not considerably change beyond this threshold.