Necessity of artificial intelligence to detect signal in the field of pharmacovigilance

Determining the benefits and possible risks of the medications helps in the rational use of medications. Documentation of the adverse events or possible risks of the medication through spontaneous reporting as a part of post-marketing safety surveillance plays a pivotal role in the pharmacovigilance program. Signal detection of previously unknown adverse events can be identified through the data mining process of big data available in large databases. These databases provide individual case safety reports that can be analyzed using statistical methods such as qualitative and quantitative approaches to measure the association between drug-event combinations, thereby confirming a signal. These approaches are time-consuming and have their own limitations such as the possibility of false safety signals due to insufficient information, over-reporting, presence of confounding factors and variation in sample size. To overcome these limitations, artificial intelligence can be essential in the early detection of signals by reducing the human efforts and time for analysis. Further advancement of quantitative methods alongside artificial intelligence can improve the process of detection of true signals.