Minimizing Spatial and Systematic Errors in Monitoring Initial Enamel Erosion

(E_t - E₀) / E₀                                                                          (E1)

E₀ and E_t being measurements at time=0 and t, respectively. To account for time-dependent systematic errors, the following has been proposed:

(E_t - R_t) / R_t                                                                          (E2)

R_t being the reference measurement at time t. However, this may introduce other errors such as those associated with spatial differences between the reference and exposed areas. Hence, we propose the following formula,

      (E_t (R₀ /R_t)- E₀) /E₀                                                                    (E3)

to correct for the systematic errors in E_t using the term (R₀ /R_t)  from the reference area.

These formulae were compared using data from a recent erosion study: half of the labial surfaces (reference) of 12 human incisors were coated with a bonding agent; and the other half were exposed to orange juice in six 10-minute intervals. At the end of each interval, the samples were rinsed with water, dried and measured with Quantitative Light-Induced Fluorescence (QLF). The exposed area was then remoistened, to recover the fluorescence, and measured again to provide a separate reference. The loss of fluorescence (dF) was measured against the initial value (E1), value from the coated area (E2 and E3), and value from the remoistened exposed area (E2). Multiple linear regression was used to relate dF with t. Results:  dF calculated with all 3 formulae demonstrated significant dose responses with erosion interval (P<0.05), but E1 gave erroneous positive values and only E3 was self-consistent (0 at time=0). E2 with the coating technique produced highest standard deviation (SD) and lowest R² value. The remoistening technique gave lower SD and higher R², but E3 gave the best results.

Conclusions: The coating technique with spatial correction minimizes systematic errors.