Patient-Reported Outcomes in Online Communications on Statins, Memory, and Cognition: Qualitative Analysis Using...

Area of Science:

• Pharmacovigilance
• Computational Linguistics
• Health Informatics

Background:

• Clinical trials face challenges in balancing participant diversity with controlled variables.
• Inclusion of underrepresented groups in trials remains a barrier, potentially masking side effects.
• Online health communities offer a supplementary data source for identifying drug side effects.

Purpose of the Study:

• To utilize a nontraditional dataset for identifying drug side-effect signals.
• To apply natural language processing (NLP) and linguistic analysis to online posts from statin users.
• To investigate the association between statin use and memory/cognitive impairment, and to capture patient-reported experiences.

Main Methods:

• Analysis of over 11 million user-generated posts from the Inspire online health community.
• Application of NLP and linguistic analysis to identify correlations between statin use and cognitive/memory issues.
• Examination of patient lexicon and themes related to statin side effects on memory.

Main Results:

• NLP identified a statistically significant correlation between statin use and discussions of memory impairment (22.6% of statin users vs. 6.2% in control).
• Linguistic analysis revealed key themes including memory loss, aphasia, cognitive impairment, and emotional changes associated with statin use.
• These findings were not observed in control groups, suggesting a specific link.

Conclusions:

• Study findings indicate a need for further research into the effects of statins on memory and cognition.
• Nontraditional datasets and NLP/linguistic methods can expand the population for identifying drug side-effect signals.
• These methods provide an additional avenue for informing patients, providers, and regulatory bodies about potential side effects.