AUTHOR=Fazmiya Mohamed Joonus Aynul , Sultana Arshiya , Heyat Md Belal Bin , Parveen Saba , Rahman Khaleequr , Akhtar Faijan , Khan Azmat Ali , Alanazi Amer M. , Ahmed Zaheer , Díez Isabel de la Torre , Ballester Julién Brito , Saripalli Tirumala Santhosh Kumar 

TITLE=Efficacy of a vaginal suppository formulation prepared with Acacia arabica (Lam.) Willd. gum and Cinnamomum camphora (L.) J. Presl. in heavy menstrual bleeding analyzed using a machine learning technique

JOURNAL=Frontiers in Pharmacology

VOLUME=Volume 15 - 2024

YEAR=2024

URL=https://www.frontiersin.org/journals/pharmacology/articles/10.3389/fphar.2024.1331622

DOI=10.3389/fphar.2024.1331622

ISSN=1663-9812

ABSTRACT=To determine the efficacy of Acacia arabica (Lam.) Willd. and Cinnamomum camphora (L.) J. Presl. vaginal suppository in Heavy Menstrual Bleeding (HMB) and on participants' health-related quality of life analyzed with machine learning algorithms. Method: A total of 62 participants in a double-dummy, single-centre study were randomly assigned to either a suppository group (SG) receiving formulation prepared with A. arabica gum (Gond Babul) and camphor from C. camphora (Kafoor) two vaginal suppositories (each weighing 3500 mg) for 7 days at bedtime along with oral placebo capsules or a tranexamic group (TG) receiving tranexamic acid (500 mg) two times a day was administered orally for 5 days and two placebo vaginal suppositories during menstruation at bedtime for three consecutive menstrual cycles. The primary outcome was the Pictorial Blood Loss Assessment Chart (PBLAC) for HMB and secondary outcomes included haemoglobin level and SF-36 HRQoL questionnaire scores. Additionally, machine learning algorithms such as k-Nearest Neighbour (KNN), AdaBoost (AB), Naive Bayes (NB), and Random Forest (RF) classifiers were employed for analysis. Results: In the SG and TG, the mean PBLAC score decreased from 635.322±504.23 to 67.70±22.37 and 512.93±283.57 to 97.96±39.25 respectively at post-intervention (TF3), demonstrating a statistically significant difference (p<0.001). A higher percentage of participants in the SG achieved normal menstrual blood loss compared to the TG (93.5% vs. 74.2%). The SG showed a considerable improvement in total SF-36 scores (73.56%) compared to the TG (65.65%) with a statistically significant difference (p<0.001). Additionally, no serious adverse events were reported in either group. Notably, machine learning algorithms, particularly AdaBoost (AB) and k-Nearest Neighbour (KNN), demonstrated the highest accuracy within cross-validation models for both primary and secondary outcomes. Conclusion: The A. arabica and C. camphora vaginal suppository are effective, cost-effective and safe in controlling HMB. This botanical vaginal suppository provides a novel and innovative alternative to traditional interventions, demonstrating promise as an effective management approach for HMB.