Sentiment Analysis of Online Reviews for Selective Serotonin Reuptake Inhibitors and Serotonin–Norepinephrine Reuptake Inhibitors

Abstract

Background: Depression affects millions worldwide, with drug therapy being the mainstay treatment. A variety of factors, including personal reviews, are involved in the success or failure of medication therapy. This study looked to characterize the sentiment of online medication reviews of Selective Serotonin Reuptake Inhibitors (SSRIs) and Serotonin–Norepinephrine Reuptake Inhibitor (SNRIs) used to treat depression. Methods: The publicly available data source used was the Drug Review Dataset from the University of California Irvine Machine Learning Repository. The dataset contained the following variables: ID, drug name, condition, review, rating, date, and usefulness count. This study utilized a sentiment analysis of free-text, online reviews via the sentimentr package. A Mann–Whitney U test was used for comparisons. Results: The average sentiment was higher in SSRIs compared to SNRIs (0.065 vs. 0.005, p < 0.001). The average sentiment was also found to be higher in high-rated reviews than in low-rated reviews (0.169 vs. −0.367, p < 0.001). Ratings were similar in the high-rated SSRI group and high-rated SNRI group (9.19 vs. 9.19). Conclusions: This study supports the use of sentiment analysis using the AFINN lexicon, as the lexicon showed a difference in sentiment between high-rated reviews from low-rated reviews. This study also found that SNRIs have more negative sentiment and lower-rated reviews than SSRIs.