Effect of Inhaled Lavender Aromatherapy on Preoperative Anxiety in Patients Undergoing Tooth Extraction: A Randomized Controlled Trial

Background/Objectives Dental anxiety remains a significant psychological barrier to oral healthcare, particularly for invasive surgical procedures such as tooth extraction. Aromatherapy using Lavandula angustifolia (lavender) has been proposed as a non-pharmacological adjunct to manage acute preoperative anxiety. This clinical trial evaluated the efficacy of inhaled lavender essential oil on anxiety severity in patients awaiting elective single-tooth extraction. Methods An assessor-blinded randomized controlled clinical trial with limited participant blinding (inherent to the recognizable scent of the intervention), registered with the Iranian Registry of Clinical Trials (IRCT20190920044824N1), was conducted among 60 patients requiring non-surgical tooth extraction. Participants were randomly assigned to an intervention group (n=30), receiving two drops of pure lavender essential oil on a sterile gauze for 20 minutes of inhalation, or a control group (n=30), receiving sweet almond oil as an olfactory placebo. The primary outcome was the severity of somatic anxiety symptoms, measured pre- and post-intervention using the Beck Anxiety Inventory (BAI). Because a significant baseline-by-treatment interaction was detected, a General Linear Model (GLM) retaining this interaction term was used to estimate adjusted group means at the grand-mean baseline, with bootstrap 95% confidence intervals (10,000 resamples). Two pre-specified checks of robustness were performed: (i) an unadjusted, baseline-naive Mann-Whitney U comparison of raw post-intervention scores, and (ii) a sensitivity analysis re-fitting the adjusted model after excluding two baseline outliers identified by inspection of model residuals. Results There were no significant baseline differences between the intervention and control groups regarding sex distribution (p=0.100), mean age (p=0.479), or pre-intervention BAI scores (Mann-Whitney p=0.215). A significant baseline-by-treatment interaction was detected (p<0.001). In the full sample (n=60), the baseline-adjusted GLM estimated a lower post-intervention BAI mean in the intervention group (22.38) than the control group (23.09), an adjusted difference of -0.71 (bootstrap 95% CI: -1.41 to -0.06; p=0.028). However, this signal was not robust: after excluding two patients identified as outliers in the model residuals (n=58), the adjusted difference attenuated to -0.51 and was no longer statistically significant (bootstrap 95% CI: -1.11 to 0.10; p=0.098). The unadjusted comparison of raw post-intervention scores was likewise not significant in the full sample (Mann-Whitney p=0.754). Conclusions A baseline-adjusted analysis of this trial produced an initial signal suggesting a small anxiolytic effect of inhaled lavender, but this signal was not stable under a pre-specified outlier-exclusion sensitivity analysis and was not supported by the unadjusted comparison of raw scores. Taken together, the totality of evidence from this trial is insufficient to conclude that inhaled lavender aromatherapy produces a reliable reduction in acute preoperative dental anxiety. Beyond its clinical findings, this trial offers a concrete, quantified illustration of how baseline imbalance and a small number of influential observations can generate a fragile, model-dependent treatment signal in a modestly sized randomized trial, a methodological caution directly relevant to the design and analysis of future small RCTs in dental and complementary medicine research.