Text Mining Analysis of Metoclopramide Reviews: Adverse Effects, Patients’ Perspectives and Off-label Use

Background: User-generated online drug reviews can give insights into patients’ perspectives on drug therapy and adverse reactions. This information can also be used to augment traditional pharmacovigilance systems that suffer from underreporting. The goal of this study was to analyze adverse effects, patients’ perspectives and offlabel uses of metoclopramide through text mining.

 

Methods: User reported drug reviews on metoclopramide were obtained from two online health forums (Drugs.com and WebMD.com) through web scraping. The raw data were preprocessed and analyzed with text mining techniques such as bag-of-words analysis and sentimental analysis. Visual data analysis techniques such as word clouds and bar plots were used to draw important conclusions from the data.

 

Results: Migraine and nausea were the most reported indications in Drugs.com and WebMD.com datasets, respectively. Text analytics show both FDA approved and off-label uses of metoclopramide. Anxiety and spasm were the most frequently reported adverse reactions in Drugs.com and WebMD.com datasets, respectively. Sentimental analysis showed that about 66% of the reviews on Drugs.com were negative, while 34% were positive. The analysis of the WebMD.com dataset revealed a similar finding with 64% negative reviews and 36% positive reviews.

 

Conclusions: User reported adverse reactions on both health forums were consistent with known adverse reactions of metoclopramide. They included both mild (drowsiness, sleepiness, dizziness, fatigue, tiredness, bloating, diarrhea) and severe adverse reactions such as movement disorders (spasm, dyskinesia, Parkinson like symptoms) and psychiatric disorders (anxiety, confusion, panic, restlessness, depression). Patients’ perspectives toward metoclopramide therapy were generally negative. Text analytics also revealed several off-label uses (migraine and hyperemesis) of metoclopramide.