Mining and evaluation of security alert signals of neuroprotective agents based on the Jinan adverse event reporting system database: a retrospective study.

BACKGROUND

Pharmacologic agents with proposed neuroprotective properties are increasingly investigated; however, limited regulation and heterogeneous evidence raise concerns about their safety profiles. To establish a foundation for its safe clinical use, the data mining technology was used to investigate the safety warning signals of neuroprotective agent post-market approval.

METHODS

The Jinan adverse event (AE) reporting system database was searched for AEs related to neuroprotective agents from January 2000 to March 2022. Basic information pertaining to patients, reports, and common AEs was analyzed. Subgroup analyses were performed to examine the distribution of adverse reaction signals for each drug across different age groups and genders. Sensitivity analyses were conducted by stratifying patients based on concomitant medication use to evaluate the distribution of adverse reaction for each drug under different stratification conditions. Kaplan-Meier curves of time were used to analyze AEs for each drug. Four disproportionality analysis methods, namely, proportional reporting ratio (PRR), reporting odds ratio (ROR), Bayesian Confidence Propagation Neural Network (BCPNN), and the Medicines and Healthcare products Regulatory Agency (MHRA), were applied to obtain alert signals for this class of drugs. We further examined the presence of the detected signals on medicine labels in China and two developed countries (USA and Japan).

RESULTS

In total, 168,314 AEs were reported, of which 2,094 were associated with neuroprotective agents. Risk signals demonstrated significant variations across age groups, gender strata, and analyses with/without concomitant medication adjustments. For the following medications-citicoline sodium, troxerutin and cerebroprotein hydrolysate, vinpocetine, leaf extract, and cerebroprotein hydrolysate injection-50% of AEs occurred before the median time point. Using the PRR, ROR, MHRA, and BCPNN methods, 81, 57, and 68 signals were detected, respectively. Among the 81 signals, 30 AEs were not included on the drug labels used in China. Of these, 11 AEs were not included on the drug labels used in Japan; two AEs were included on the drug labels used in China but not in Japan.

CONCLUSION

The Jinan AE reporting system database used to mine warning signals can be used to analyze AEs after the marketing of neuroprotective agents, thereby reducing the risk associated with their clinical use.