Health gain from restorative dental treatment evaluated by computer simulation

To simulate the influence over time of relevant factors on health gain from restorative dental treatment under varying assumptions and to compare outcomes with those resulting under a 'do nothing' scenario.

A decision model was used in a computer simulation of the caries process in posterior approximal tooth surfaces. The effect of superimposing restorative treatment, based on bitewing radiology, was incorporated. Input data came from current observational studies and included caries progression rates, notional survival times of restorations, and sensitivity (Sn) and specificity (Sp) of treatment decision making by a high (A) and a low (B) performing dentist. Utility values, on a scale 0-1, for different tooth health states were obtained from questioning 92 adults of appropriate age.

By assumption, the primary dental care sector.

A hypothetical population, initially 14-15 years old.

Class II amalgam restorations.

Computed in utility based units (UBUs) as sums of the product of numbers of sound, carious and restored surfaces and their utility values. Health gain (in UBUs) was assessed relative to interim end point UBUs pertaining under 'do nothing'.

One thousand approximal surfaces, designated initially as 920 sound, 51 carious and 29 filled were followed in the model over 10 years. The greatest health gain (33.16 UBUs) was from dentist A (Sn = 0.23, Sp = 0.99, 50 per cent restoration survival rate = 10 years, caries rate = 4.4% per annum). The least was from dentist B (Sn = 0.52, Sp = 0.88, 50% survival = 5 years, caries rate = 0.0% per annum) representing a loss of 17.07 UBUs compared with 'do nothing'.

Results suggest that caution is advisable when making positive decisions to restore on the basis of bitewing radiographs.