FAERS is a publicly accessible database that aggregates and summarizes adverse drug reaction reports from across the globe, providing a comprehensive view of ADE occurrences and facilitating post-marketing safety monitoring. It houses over 9 million individual drug-related adverse event reports submitted by industry professionals, physicians, pharmacists, healthcare professionals, consumers, and others, making it the largest spontaneous reporting system database worldwide (Okada et al., 2019; van Hasselt et al., 2020). The FAERS data is categorized into seven datasets: demographic and administrative information (DEMO), drug information (DRUG), adverse drug reaction information (REAC), patient outcomes information (OUCT), reported sources (RPSR), drug therapy start dates and end dates (THER), and indications for drug administration (INDI). A relationship is established within the FAERS database architecture that connects each data file through a unique identification number (Zhou et al., 2024). Detailed data from FAERS can be downloaded from the FDA website (http://www.fda.gov/).

The JADER database contains information on cases reported by pharmaceutical companies and medical institutions since 2004. The JADER database comprises four files: DEMO, DRUG, REAC, and HIST (Imai et al., 2022). The “DEMO” file includes basic patient information such as sex, age, and weight. The “DRUG” file contains the generic name of the drug, route of administration, and the start and end dates of administration. The “REAC” file records the name of the adverse event, outcome, and the date of occurrence. The “HIST” file contains information about the patient’s underlying condition (Inaba et al., 2019). Data from JADER were downloaded from the Pharmaceutical and Medical Devices Agency website (https://www.pmda.go.jp/index.html).

In our study, we utilized two spontaneous reporting system databases, FAERS and JADER, to mitigate reporting bias and enhance the reliability of the results. Considering the differing marketing times of irbesartan, the analysis period for this study spans from April 2004 to March 2024 in FAERS and from April 2008 to December 2023 in JADER. In FAERS, searches were conducted using generic and brand names such as “IRBESARTAN,” “APROVEL,” and “KARVEA.” In JADER, “イルベサルタン” was used for retrieval. Since FAERS updates quarterly, which may encompass duplicate reports or those that have been withdrawn or deleted, we performed deduplication following FDA recommendations to address duplicate reports submitted by different sources: In the DEMO file, we selected the PRIMARYIDs, CASEIDs, and FDA_DTs, subsequently sorting them by CASEIDs, FDA_DTs, and PRIMARYIDs. (1) If CASEIDs are the same, the latest FDA_DTs is selected; (2) If CASEIDs and FDA_DTs are the same, the higher PRIMARYIDs is selected (Wan et al., 2022). ADE reports associated with irbesartan use were extracted, standardized, and mapped to Medical Dictionary for Regulatory Activities (MedDRA 26.0) (Singh, 2015). The role code for ADEs was retained as the primary suspect (PS) by mitigating drug interactions (I), concomitant drugs (C), secondary suspect drugs, and other unknown drugs that may cause ADEs (Ji et al., 2022; Zhang et al., 2023). Based on MedDRA’s structural hierarchy, the screened ADEs were mapped into preferred terms (PTs) and system organ classes (SOCs). In FAERS, we conducted a statistical analysis of ADE reports, focusing on variables such as sex, age, weight, indication, reporting countries, outcome, and the reported person. During subgroup analyses, we excluded reports with missing data for sex and age. In the JADER database, we removed duplicate data from the DRUG and REAC files and linked the DEMO table to these tables based on the identified conditions (Kato et al., 2022). The drugs were classified into three categories—“PS,” “C,” and “I”—according to their impact on adverse events. Consistent with our approach in FAERS, we designated the role code as PS. Adverse events were coded following the terminology recommended in MedDRA, Japanese version 26.0 (Yamaoka et al., 2022). It is important to note that the DEMO file includes demographic information organized in 10-year age intervals (e.g., 60–69 years). We excluded missing sex data and non-numeric age categories, such as early, mid, late gestation, neonatal, infant, pediatric, young adult, adult, elderly, and unknown, from the subgroup analyses (Kose et al., 2023).

Considering the differences in reporting formats and data structures between FAERS and JADER, we employed the following methods to ensure the comparability of the analysis results:1: Uniform coding system: We utilized a standardized coding system for medical terminology (MedDRA 26.0) to mitigate bias in data analysis stemming from the differing database formats and to ensure the comparability of terms used for adverse reactions across both databases (Zou et al., 2024). 2: Removal of duplicate and erroneous records: We eliminated spontaneous reporting of duplicate or erroneous records within the databases to reduce inconsistencies and to enhance the completeness and credibility of the data analyses. 3: Bayesian algorithms: We employed Bayesian-based algorithms and empirical Bayesian geometric averaging, which effectively address variations in data by adapting to different distributions and structures, thereby generating consistent and reliable signals to ensure comparability of results (Ewers et al., 2012; Agrawal et al., 2019). 4: Timeframe selection: We selected the period following the FDA approval of Irbesartan (September 1997) and its approval in Japan (April 2008) for analysis to minimize bias associated with reporting during the clinical trial phase.

For pharmacovigilance, disproportionality analysis is commonly employed with post-marketing surveillance databases to assess associations between drugs and adverse events (Shen et al., 2019). This analysis compares the observed and expected number of reports for any given combination of drug and adverse event, generating hypotheses about possible causal relationships. In our study, we applied both Bayesian and frequency (non-Bayesian) methods to evaluate the relationship between irbesartan and ADEs. Frequency methods include Reporting Odds Ratio (ROR) (van Puijenbroek et al., 2002) and Proportional Reporting Ratio (PRR) (Evans et al., 2001). Bayesian methods encompass the Bayesian Confidence Propagation Neural Network (BCPNN) (Bate et al., 1998) and the Multi-Item Gamma Poisson Shrinker (MGPS) (Szarfman et al., 2002). While frequency methods are computationally simple and highly sensitive, they are prone to false positives when the number of adverse events is small (Wu et al., 2023). In contrast, Bayesian methods account for uncertainty in the disproportionality rate when reported cases are limited (Almenoff et al., 2007). In particular, the ROR is more suitable for high-frequency adverse event reporting, offering the advantage of correcting for bias due to a low number of reports for certain events (Fan et al., 2024). In contrast, the PRR is recognized for its higher specificity compared to ROR (Jiang et al., 2024). The BCPNN excels in integrating data from multiple sources and demonstrates robust performance in cross-validation, which helps to reduce the occurrence of false-positive signals. However, it tends to be more conservative, potentially missing signals for very rare events (Lin et al., 2024). In this regard, the MGPS algorithm offers a more comprehensive approach, particularly effective in detecting signals associated with rare events (Li Y. et al., 2024). Each method contributes uniquely to the detection of drug-associated safety signals, and their combined use provides a more balanced and comprehensive understanding of potential drug-adverse event associations (Zou et al., 2023) (Table 1). We define a PT as a positive signal if it simultaneously meets the thresholds of all four methods (Cui et al., 2023). To mitigate the risk of false-positive results (type I errors), we employed the Bonferroni method to adjust for multiple comparisons of P-values (Curtin and Schulz, 1998). The adjusted threshold is computed as: Bonferroni-corrected P-value = P/n, where P is the original significance threshold and n represents the total number of tests conducted. R software (version 4.2.1) was used for data processing and statistical analysis.

The time to onset (TTO) of irbesartan-related ADEs was defined as the interval between EVENT_DT (date of ADE onset in the DEMO file) and START_DT (date of medication initiation in the THER file). Cases with missing dates (either the initiation of medication or the onset of ADEs) or inaccuracies (not specified to a particular day, month, or year) were excluded. Additionally, cases in which the onset date of ADE occurred before the initiation date of irbesartan therapy were also excluded, as this would result in a negative time-to-onset calculation (Zou et al., 2024). We employed the median, quartile, minimum, maximum, and Weibull shape parameter to comprehensively assess TTO characteristics in our study (Kinoshita et al., 2020). The Kaplan-Meier method was employed to illustrate the cumulative incidence of ADEs associated with irbesartan (Wang et al., 2024). The frequency of adverse events following the initiation of treatment is contingent upon the drug’s mechanism of action and may vary over time. In contrast, the incidence of adverse events not associated with drug therapy tends to be more stable (Cornelius et al., 2012). Changes in the risk incidence of ADEs over time can be identified and predicted by the Weibull distribution test, with scale (α) and shape (β) parameters characterizing the Weibull distribution’s shape (Mazhar et al., 2021). In this study, we focus exclusively on the parameter β. When the shape parameter β is less than 1 and its 95% confidence interval (CI) is also below 1, the risk of adverse effects is considered to decrease over time, indicative of an early failure-type curve. Conversely, if β is approximately equal to 1, with its 95% CI containing the value of 1, the risk is estimated to remain persistent over time, representing a random failure-type curve. Finally, if β is greater than 1 and its 95% CI does not encompass the value of 1, the hazard is interpreted as increasing over time, characteristic of a wear-out failure-type curve (Mazhar et al., 2021; Sauzet et al., 2013).

A total of 5,816 and 366 ADE reports were obtained from the FAERS (January 2004-January 2024) and JADER (April 2008-December 2023) databases, respectively (Figure 1). The number of ADE reports in the FAERS database remained above 400 annually from 2018 through 2023 (Figure 2A), whereas in JADER, the number of reports did not exceed 50 per year (Figure 2B).

The characteristics of ADEs reported in both databases, including age, weight, sex, outcome, indication, and reported country, are detailed in Tables 2, 3. In FAERS, ADEs were more frequently reported in female patients (n = 2,916, 50.1%) than male patients (n = 2,095, 36.0%). Among the known age reports, patients aged 65 years or older constituted a major portion (n = 2,878, 49.5%). Body weight, a critical indicator for evaluating adverse reactions to chemical drugs (Yuan et al., 2020), was missing for most patients (n = 4,075, 70.1%). Most reports were submitted by health professionals (n = 3,707, 63.8%), with serious adverse outcomes mainly being other-serious medical events (n = 2,797, 38.2%) and hospitalization-initial or prolonged (n = 2,682, 36.6%). Notably, the proportion of deaths and life-threatening events reported was 3.6% and 5.5%, respectively. The primary indication for irbesartan use was hypertension (n = 2,997, 51.5%), with other indications including essential hypertension (n = 74, 1.2%) and cardiac failure (n = 57, 0.9%) (Table 2).

Conversely, in JADER, higher proportions of male patients (n = 204, 55.7%) submitted more ADE reports compared to female patients (n = 152, 41.5%). The number of reported cases for individuals under 65 years of age is 135, accounting for 36.9% of the total reports. In contrast, the number of reports for individuals aged 65 years or older is 212, representing 57.9% of the total. Among those who provided specific body weight information, the largest number of reports were in the 50–100 kg weight range (n = 158, 43.2%). Sixty nine percent of the submitters experienced varying degrees of recovery or rehabilitation, but 5.2% of the submitters died. Consistent with FAERS reports, the primary indication for irbesartan was hypertension (n = 291, 72.6%) (Table 3).

Mapping the PTs in ADE reports to the corresponding SOC level, we counted the number of reports in both databases. We ranked the case number of PTs in descending order for each SOC and found that irbesartan-associated PTs mainly involved 27 SOCs in FAERS and 22 SOCs in JADER. The top 3 SOCs in terms of reported cases differed between the databases. In FAERS, they were general disorders and administration site conditions (n = 2,186), nervous system disorders (n = 1,881), and gastrointestinal disorders (n = 1,559) (Figure 2C). In JADER, they were metabolism and nutrition disorders (n = 94), cardiac disorders (n = 61), and investigations (n = 58) (Figure 2D). However, among the top 5 SOCs, three overlapped: nervous system disorders, metabolism and nutrition disorders, and investigations.

The signal strength of ADE reports in both databases at the SOC level was calculated using disproportionality analysis. Based on the ROR method, we plotted the forest of signal strength. In FAERS, ten SOCs met the ROR positive threshold: nervous system disorders (SOC code: 10029205), investigations (SOC code: 10022891), metabolism and nutrition disorders (SOC code: 10027433), renal and urinary disorders (SOC code: 10038359), respiratory, thoracic and mediastinal disorders (SOC code: 10038738), vascular disorders (SOC code: 10047065), cardiac disorders (SOC code: 10007541), hepatobiliary disorders (SOC code: 10019805), ear and labyrinth disorders (SOC code: 10013993), and endocrine disorders (SOC code: 10014698) (Figure 2E). In JADER, four SOCs met the positive threshold: metabolism and nutrition disorders, cardiac disorders, renal and urinary disorders, and musculoskeletal and connective tissue disorders (SOC code: 10028395) (Figure 2F). The three SOCs positive in both databases were metabolism and nutrition disorders, cardiac disorders, and renal and urinary disorders.

Notably, in FAERS, the signal strength of renal and urinary disorders (ROR: 3.33 [3.13–3.53], PRR: 3.19, EBGM05: 3.00, IC025: 0.01) met the positive threshold for all four disproportionality methods (Table 4). In JADER, metabolism and nutrition disorders (ROR: 5.17 [4.14–6.46], PRR: 4.45, EBGM05: 3.55, IC025: 0.48) was the only SOC that simultaneously met the positivity threshold for all four methods (Table 5). The complete results are presented in Tables 4, 5.

Subsequently, considering only the frequency of reports, we identified the top 20 PTs in both cohorts. In FAERS, the PT with the highest number of reported cases was acute kidney injury (n = 586, 3.18%), followed by hyponatremia (n = 343, 1.86%), hypotension (n = 326, 1.77%), drug ineffective (n = 321, 1.74%), and dizziness (n = 272, 1.48%) (Figure 3A). In contrast, the top 5 PTs in the JADER database were hyperkalemia (n = 52, 9.65%), rhabdomyolysis (n = 22, 4.08%), renal impairment (n = 19, 3.53%), hepatic function abnormal (n = 18, 3.34%), and interstitial lung disease (n = 14, 2.60%) (Figure 3B). Additionally, we identified six overlapping signals: acute kidney injury, hyponatremia, hypotension, hyperkalemia, diarrhea, and bradycardia.

Using disproportionality methods, we calculated the signal strength of each PT and filtered out all signals that simultaneously met the positivity thresholds. In FAERS, 219 signals met the criteria, while in JADER, 20 signals were identified. Based on the descending order of filtered signals from the reported cases and grouped according to SOC, we presented the ROR values and their 95% CIs for the top 20 signals in each cohort using a forest plot. Generally, higher ROR values indicate a stronger association of these signals with irbesartan. In FAERS, several PTs had a high number of reports along with relatively strong signal strength, including hyponatremia (n = 343, ROR: 20.68 [18.58–23.02], PRR: 20.31, EBGM05: 18.45, IC025: 2.67), hyperkalemia (n = 234, ROR: 23.16 [20.35–26.36], PRR: 22.88, EBGM05: 20.38, IC025: 2.84), and orthostatic hypotension (n = 74, ROR: 14.04 [11.17–17.65], PRR: 13.99, EBGM05: 11.50, IC025: 2.13). Despite the small number of cases, some signals demonstrated strong signal strength, including amyloid arthropathy (n = 11, ROR: 4361.81 [728.77–1602.36], PRR: 820.04, EBGM05: 365.48, IC025: 7.61), neurologic neglect syndrome (n = 22, ROR: 156.19 [101.69–239.91], PRR: 156.01, EBGM05: 148.11, IC025: 5.54), carotid artery thrombosis (n = 24, ROR: 108.66 [72.27–163.36], PRR: 108.52, EBGM05: 74.40, IC025: 5.04), and personality disorder (n = 24, ROR: 30.10 [20.13–45.01], PRR: 30.06, EBGM05: 21.26, IC025: 3.23) (Supplementary Table S1). Notably, we identified several new signals not listed in the drug label, including hyponatremia, syncope, lactic acidosis, arrhythmia, acute pancreatitis, rhabdomyolysis, and cholestasis (Figure 4A). Supplementary Table S1 provides the full results of the analysis in FAERS.

In the JADER database, the filtered 20 signals are listed in Figure 4B. PTs with a high number of reported cases included hyperkalemia (n = 52, ROR: 39.26 [29.44–52.34], PRR: 35.57, EBGM05: 27.60, IC025: 3.46), rhabdomyolysis (n = 22, ROR: 8.02 [5.23–12.29], PRR: 7.73, EBGM05: 5.39, IC025: 1.28), and renal impairment (n = 19, ROR: 3.46 [2.19–5.47], PRR: 3.38, EBGM05: 2.30, IC025: 0.08). PTs with strong signal strengths included hyperkalemia, increased blood potassium (n = 9, ROR: 33.70 [17.37–65.39], PRR: 33.15, EBGM05: 18.81, IC025: 3.36), procedural hypotension (n = 4, ROR: 125.32 [45.82–342.73], PRR: 124.39, EBGM05: 51.24, IC025: 5.20), and hypochloremia (n = 4, ROR: 131.43 [48.00–359.84], PRR: 130.46, EBGM05: 53.57, IC025: 5.26). New signals included cardiac failure (n = 12, ROR: 3.75 [2.12–6.65], PRR: 3.69, EBGM05: 2.28, IC025: 0.21), hyponatremia (n = 11, ROR: 7.53 [4.14–13.69], PRR: 7.40, EBGM05: 4.47, IC025: 1.21), respiratory failure (n = 7, ROR: 6.76 [3.20–14.26], PRR: 6.68, EBGM05: 3.57, IC025: 1.07), tubulointerstitial nephritis (n = 6, ROR: 5.77 [2.58–12.92], PRR: 5.72, EBGM05: 2.91, IC025: 0.84), and blood urea increased (n = 5, ROR: 13.34 [5.52–32.25], PRR: 13.23, EBGM05: 6.29, IC025: 2.05) (Figure 4B). Supplementary Table S2 provides the full results of the analysis in JADER.

Combining the results from both databases, we identified eight positive new signals screened in both databases: hyponatremia, hypotension, hyperkalemia, bradycardia, angioedema, rhabdomyolysis, tubulointerstitial nephritis, and hypochloremia (Figure 4C). To visually represent the most significant ADE signals, volcano plots were generated for the analysis results of both FAERS (Supplementary Figure S1A) and JADER (Supplementary Figure S1B) databases. The plots display 219 and 20 positive signals, respectively. In these plots, the horizontal axis represents the log2-transformed ROR values, while the vertical axis shows the -log10-transformed corrected P-values (adjusted using the Bonferroni method). Signals on the right side of the plot, corresponding to higher log2-transformed ROR values, indicate a stronger association with irbesartan compared to those on the left. Strong positive signals identified in FAERS include rare events such as atrial standstill and amyloid arthropathy. In the JADER database, notable signals include hypochloremia and procedural hypotension. These signals represent adverse drug events that show a strong association with irbesartan based on the data from these pharmacovigilance databases, further highlighting the importance of monitoring specific reactions in different populations.

To minimize the influence of confounding factors, we conducted subgroup analyses of adverse events associated with irbesartan using data from both the FAERS and JADER databases. The top 15 most common adverse events for each subgroup were identified based on positive signal criteria (Supplementary Figures S2A, C, E, G). In FAERS, rhabdomyolysis and acute pancreatitis were male-specific, while aphasia and arrhythmia were female-specific (Supplementary Figure S2B). Signals such as muscle spasticity, urinary retention, and dysphagia were predominant in patients under 65, while bradycardia, eczema, and orthostatic hypotension were observed in those 65 and older (Supplementary Figure S2D). In JADER, four overlapping signals—hyperkalemia, rhabdomyolysis, blood potassium increased, and blood creatinine increased were found in both male and female subgroups (Supplementary Figure S2F). Age-specific signals included renal impairment and hepatic function abnormal in the age less than 65 subgroup, and agranulocytosis and respiratory failure in those aged 65 and older (Supplementary Figure S2H).

In clinical practice, irbesartan is often co-administered with other antihypertensive agents such as hydrochlorothiazide and amlodipine to enhance blood pressure control. To eliminate the potential effects of concomitant medications on our results, we conducted a sensitivity analysis. After excluding cases with co-administration of other drugs, we identified 1,479 reports. Persistent adverse reactions included hyponatraemia, hyperkalemia, angioedema, acute pancreatitis, arrhythmia, increased blood creatine phosphokinase, presyncope, swollen tongue, and agranulocytosis, among others (Supplementary Table S3).

Due to the limited number of effective TTO reports in the JADER database, we only statistically analyzed the TTO reports in the FAERS database. In FAERS, there were 987 (17.0%) total effective TTO reports. The median onset time of all ADEs was 107 days, with an interquartile range (IQR) of 15–469 days (Figure 5B).

ADEs following irbesartan administration occurred primarily within 1 month of administration (n = 337, 34.14%) and after 1 year of administration (n = 319, 32.32%). Within 180 days, the number of TTO reports tended to decrease over time, but reports beyond 180 days still accounted for about 40% of cases (Figure 5A). The Weibull distribution test for TTO indicated that the upper limit of the 95% confidence interval (CI) for the shape parameter (β) was less than 1 in FAERS (0.55), indicating an early failure type, suggesting that the probability of an ADE gradually decreased over time (Figure 5B).

Additionally, we analyzed the TTO reports at the SOC level. Among the 23 SOC levels with at least 10 valid TTO reports, there was a significant difference in TTOs (P < 0.0001, Figure 5C). SOCs with the shortest median onset times included “product issues” [median onset time (MOT): 10 days], “immune system disorders” (MOT: 17 days), and “eye disorders” (MOT: 31 days). SOCs with the longest median TTOs included “surgical and medical procedures” (MOT: 3,667 days), “neoplasms benign, malignant, and unspecified (incl. cysts and polyps)” (MOT: 911 days), and “injury, poisoning, and procedural complications” (MOT: 730 days) (Supplementary Table S4). The cumulative incidence of ADEs over time is depicted in a Kaplan-Meier plot (Figure 5D).

Our analysis of baseline information revealed notable differences in the demographic data concerning irbesartan ADEs between the FAERS and JADER databases. In FAERS, the majority of ADE reports were submitted by females (50.1%) compared to males (36.0%), with 49.5% of submitters aged over 65 years. Despite a significant portion of reports lacking specific weight information, weights over 100 kg represented a major segment of the known weights. The primary reporting regions were the Americas (United States, Canada) and Europe (France, United Kingdom, and Italy), constituting 61.7% of the reports.

Conversely, in JADER, the proportion of male submitters (55.7%) exceeded that of females (41.5%). While 57.9% of the submitters were aged 65 years or older, 36.9% were under 65 years of age, and weights fell primarily within the 50–100 kg range (43.2%), with the vast majority of reports originating from Japan. These demographic variations can be partially attributed to the epidemiological characteristics of hypertension. In 2016, 34.6% of Japanese men and 24.8% of Japanese women were reported to suffer from hypertension (Asakura et al., 2021). By 2017, the prevalence of hypertension among the Japanese working-age population (20–64 years) was 37.5%. Notably, mean systolic and DBP levels have decreased significantly in middle-aged and older Japanese adults due to advancements in hypertension treatment (Hisamatsu et al., 2020). Examining trends over a 55-year period from 1961 to 2016, DBP levels in women decreased by 4–8 mm Hg across all age groups. However, in men aged 50–59, DBP levels remained unchanged, and in men aged 30–49, they increased, potentially due to rising obesity rates, reduced physical activity, and inadequate diastolic hypertension treatment (Nagai et al., 2015; Hisamatsu and Miura, 2024). Similarly, a report from the American Heart Association indicates that in the United States, the prevalence of hypertension is increasing twice as fast in women compared to men, with older women (≥65 years) showing higher prevalence rates than men (Mozaffarian et al., 2015). This trend may be related to post-menopausal hormonal changes, activation of the RAS system, sympathetic nervous system, and increased anxiety and depression levels (Yanes and Reckelhoff, 2011; Maric-Bilkan and Galis, 2016).

The epidemiological studies mentioned above support our analysis. It is noteworthy that differences reported by ADEs in the two databases could be attributed to lifestyle and genetic (ethnic) differences between Japan and Western countries. Hypertension, a multifactorial disease influenced by environmental and genetic factors, shows differing prevalence and control rates across regions (Takeuchi et al., 2018). Japan has seen relatively lower improvements in hypertension awareness, treatment, and control compared to the United States and Europe (Sekikawa and Hayakawa, 2004). Despite higher obesity rates in European and American populations, East Asians are genetically more sensitive to salt and consume higher amounts of it (Bailly et al., 2020; Alfaras et al., 2016). Additionally, Japanese men’s lifestyle choices, such as excessive alcohol consumption and smoking, exacerbate hypertension (Kokubo, 2014). Interestingly, the indications for irbesartan were highly consistent across the two databases. Thorough analysis of these baseline data differences is crucial, as these demographic characteristics may introduce bias into ADE outcomes. Understanding these variations can enhance the interpretation of cohort differences and improve the accuracy of our findings.

In drug safety evaluation, it is crucial to assess the interval between drug administration and the onset of ADEs. This assessment can elucidate the underlying mechanisms of ADEs, identify specific time windows of risk during treatment, and facilitate earlier prevention or diagnosis of adverse reactions (Leroy et al., 2014). The timing of ADEs following RAS inhibitor use has been previously reported. For instance, angioedema after ACEI use typically occurs within the first week of treatment, although a significant proportion can manifest after months or even years (Howes and Tran, 2002). Diseases such as pemphigus associated with ARB therapy often appear months or years after treatment initiation (Steckelings et al., 2001).

A study by Zahedi I et al. suggests that liver function enzymes should be monitored in high-risk patients from the initiation of losartan, from a few days to several months (Zahedi et al., 2023). In our study, the median TTO of “hepatobiliary disorders” and “skin and subcutaneous tissue disorders” was 195 and 46 days, respectively, aligning with the clinical studies mentioned. Additionally, we found that the adverse effects of irbesartan predominantly occurred at 1 month (34.14%) and 1 year (32.32%) after administration. Acute hypersensitivity or immune-mediated reactions, such as angioedema or anaphylactic rash, may occur early in irbesartan therapy, supported by median onset times of 47 days for “skin and subcutaneous tissue disorders” and 17 days for “immune system disorders” in our TTO analysis. Early drug interactions, especially with concurrent medications, may lead to increased irbesartan concentrations, causing adverse reactions like hypotension and dizziness (Nunnery and Mayer, 2019). Long-term use can gradually impair renal function, particularly in patients with chronic kidney disease, as indicated by a median onset of 161 days for “renal and urinary disorders”. Furthermore, long-term therapy may result in drug tolerance or adaptation, leading to an increased frequency of adverse reactions over time (Brass, 1984). These findings highlight the need for clinicians to closely monitor patients in the early stages of treatment to promptly detect immune reactions or organ damage. Additionally, regular assessments of renal and hepatic function, as well as electrolyte levels, are essential during long-term therapy, especially beyond 1 year, to prevent the progression of chronic adverse reactions. In summary, our findings underscore the necessity for continuous vigilance in clinical practice (Maignen et al., 2010).

It is important to acknowledge the limitations of our study:

1. Differences in case reports: The total number of cases in the two databases varied significantly; JADER was limited to case reports, whereas FAERS included periodic reports that encompassed non-serious cases over an extended period. This discrepancy may have influenced the results of our analyses (Zou et al., 2024).

Despite these limitations, the combined FAERS and JADER databases provide valuable resources for post-marketing safety monitoring of irbesartan.