Objectives: The purpose is to contrast the time-series changes in vital signs, cost-effectiveness, sedation depth, and patient experience between midazolam and dexmedetomidine during third molar extraction retrospectively.
Materials and methods: This...
Objectives: The purpose is to contrast the time-series changes in vital signs, cost-effectiveness, sedation depth, and patient experience between midazolam and dexmedetomidine during third molar extraction retrospectively.
Materials and methods: This retrospective study involved patients who underwent conscious sedation with midazolam or dexmedetomidine during third molar extraction. Outcome variables included vital signs (systolic blood pressure(SBP), diastolic blood pressure(DBP) and heart rate(HR)) and bispectral index (BIS) scores recorded at surgical time points (before sedative administration [T0], right after sedative administration [T1], during local anesthesia [T2], 20 minutes post local anesthesia [T3], and at primary closure [T4]), drug cost, Observer Assessment of Alertness/Sedation Scale (OAA/S), amnesia(0-3), satisfaction(1-5). Covariates collected for adjustment included sex, age, body mass index (BMI), sleep time ratio, Dental Anxiety Score (DAS), and Pederson scale for extraction difficulty.
Categorical variables were compared using chi-square tests, while continuous variables were compared using independent t-tests for basic statistical analysis to evaluate baseline physiological differences between groups. Subsequently, propensity score matching (1:1, caliper=0.2) was performed to correct baseline imbalances between groups.
Vital sign volatility was analyzed using four indices: maximum percentage change (max % change), standard deviation (SD), coefficient of variation (CV), and area under the curve of absolute percentage change (AUC₍abs₎). The maximum % change represents the highest absolute percentage change observed at any surgical time point for each patient, indicating the greatest fluctuation. The standard deviation reflects the dispersion of percentage changes over time, representing overall volatility. The coefficient of variation(CV) was calculated as the standard deviation(SD) divided by the absolute mean of the percentage changes, indicating relative volatility. AUC₍abs₎ is area under the curve of absolute percentage change, defined as the total cumulative change during surgery, was calculated as the time-integrated area under the absolute percentage change curve over T1–T4. Differences in group means were evaluated with independent t-tests. For heart rate, which showed significant differences among these indices, linear mixed-effects models (LMM) were applied to evaluate interactions among sedative agent, time, and covariates while multiple linear regression models (LM) analysis was also used to identify risk factors for maximum heart rate % change. The risk factors contributing to outlier 20 were examined using logistic regression, which were defined as event occurring at one or more time points with an absolute HR percentage change ≥20% relative to T0
Cost and effect analyses were conducted using three indicators. E1 (Outlier avoidance) represents the proportion of cases in which neither a heart rate change of 20% or more nor clinical instability (OAAS <3 or BIS <70) occurred at any surgical time point. E2 (Composite stability) is the average z-score of the maximum percentage changes in systolic/diastolic blood pressure and heart rate. E3 (Composite success) is the proportion of cases meeting both criteria of adequate sedation (OAAS = 3 or 4, BIS between 80 and 85) and high patient satisfaction (score of 4 or 5). Additionally, incremental cost-effectiveness ratios (ICERs) were computed for each indicator based on costs by group.
A linear mixed-effects model (LMM) was used to analyze BIS volatility over time, and Pearson correlation was employed to assess the association between OAA/S and BIS values. Ordinal logistic regression was performed to assess differences in patient-reported measures (amnesia, satisfaction) by drug, and a cumulative ordinal logistic model was fitted to compare drug responses by quantiles of DAS, pederson, sleep time ratio.
Results: After propensity score matching, standardized mean differences (SMDs) for age, sex, sleep time ratio, DAS, Pederson, BMI, vital signs and BIS at T0 time point were all below 0.2, indicating balanced groups with 66 patients per group. In the vital sign volatility indices, no significant differences were found in systolic/diastolic blood pressure; however, the maximum heart rate change was significantly higher in the MDZ group compared to the DEX group (MDZ 37.99 ± 21.99% vs. DEX 26.31 ± 21.11%, p=0.010*). The standard deviation (SD) was also lower in the DEX group (MDZ 18.07 ± 9.29 vs. DEX 12.70 ± 9.16, p<0.001***), while the coefficient of variation (CV) was relatively higher in the DEX group (DEX 0.97 ± 0.34 vs. MDZ 1.10 ± 0.34, p=0.030*). Furthermore, the area under the curve of absolute changes (AUC₍abs₎) was significantly lower in the DEX group (MDZ 699.62 ± 478.59 vs. DEX 406.04 ± 396.63, p<0.001***).
Linear mixed-effects modeling further revealed that the increase in heart rate was markedly attenuated over time in the DEX group, indicating overall more stable heart rate responses, whereas the MDZ group exhibited progressively increasing heart rate up to 20 minutes after local anesthesia. Multiple linear regression analysis of maximum heart rate changes showed a significant reduction in peak heart rate amplitude during surgery by an average of 12.93% points in the DEX group versus the MDZ group. Female sex and younger age were associated with significantly greater heart rate volatility (p<0.05*), and higher heart rate at T0 was linked to a smaller relative maximum change (p<0.000***). In logistic regression for risk factors of outlier 20, sedatives showed a statistically significant association with outlier occurrence risk. The DEX group had significantly lower odds of outlier 20 compared to the MDZ group, with an adjusted odds ratio (aOR) of 0.11 (95% CI 0.04–0.32, p<0.001***). Age yielded aOR=0.91 (95% CI 0.85–0.98, p=0.009*) and HR at T0 had aOR=0.96 (95% CI 0.93–0.99, p=0.013*).
In terms of drug cost, the average cost for the MDZ group was 0.373 ± 0.045 USD, whereas the DEX group incurred a cost of 29.27 USD (SD 0). Regarding effect indicators, the E1 (Outlier avoidance) was high in both groups, 0.955 in the MDZ group and 1.000 in the DEX group. The E2 (Composite stability) was -0.117 for the MDZ group and 0.117 for the DEX group, while the E3 (Composite success) was 0.136 and 0.030, respectively. The incremental cost-effectiveness ratio (ICER) for E1 was 635.74 USD per outlier avoided, with a confidence interval (CI) ranging from 272.46 USD to 1,907.55 USD and a median of 635.74 USD. For E2, the ICER was 123.39 USD per one-point improvement in stability, with a CI of 52.34 USD to 700.71 USD and a median of 121.26 USD. The ICER for E3 was -272.46 USD per one unit of success, with a CI ranging from -953.86 USD to -136.24 USD and a median of -272.48 USD.
The linear mixed-effects model (LMM) analysis of BIS changes revealed a significant initial decline after sedative administration, followed by gradual recovery over time. Higher predicted BIS values were observed in the DEX group compared to the MDZ group at time points T1 and T2. No significant differences were noted at time points T3 and T4, indicating a convergence of sedation levels over time. Pearson correlation analysis between BIS at T4 and OAA/S showed correlation coefficients of r=0.203 (p=0.019*) and r=0.352 (p<0.001***) for BIS_min. Partial correlation analyses controlling for covariates such as sex, age, BMI, sleep time ratio, DAS, pederson, and BIS at T0 demonstrated similar positive correlations between OAA/S and BIS at T4 (r=0.200, p=0.026*) and BIS_min (r=0.341, p<0.001***)
Ordinal logistic regression analysis for amnesia levels (0–3) showed an odds ratio of 0.93 (95% CI 0.49–1.77, p=0.828), indicating no statistically significant difference between groups. For satisfaction levels (1–5), the odds ratio was 0.62 (95% confidence interval 0.31–1.23, p=0.178), also showing no significant difference. Cumulative logit ordinal logistic regression by quantiles (25%, 50%, 75%) of DAS, sleep time ratio, and pederson indicated no significant difference in amnesia or satisfaction between the two drugs at any stratification level, with adjusted odds ratios for all below 1 and not statistically significant.
Overall, These results suggest that dexmedetomidine demonstrates superior heart rate stability compared to midazolam. However, midazolam shows overall superiority in cost-effectiveness. Both sedatives exhibit similar effects in patient-reported measures.
Conclusion: When selecting dexmedetomidine over midazolam for third molar extraction, superior heart rate stability, relatively higher cost, and lighter sedation level in early stage should be the primary considerations rather than patient-reported outcomes
Keyword: Dexmedetomidine, Midazolam, Vital sign, Volatility indices, Cost-effectiveness, ICER, Bispectral Index(BIS) value, OAA/S (Observer’s Assessment of Alertness/Sedation Scale), DAS(Dental Anxiety Score), Multiple linear regression, Linear mixed model, Logistic regression