The benefits of medication therapy for patients are evident in preventing and treating various diseases, as well as improving or maintaining the quality of life (1). However, individuals with multimorbidity often experience poor functional status and health outcomes, and are frequently treated by multiple healthcare specialists, resulting in frequent hospital visits, polypharmacy, and a substantial treatment burden (2–4). Polypharmacy is associated with a significantly increased risk of drug-drug interactions (DDIs) and adverse drug events (ADEs) among patients with multiple comorbidities (5, 6).

As one of the preventable drug-related problems, DDIs occur when a patient is simultaneously exposed to two or more medications that are known to interact. These interactions might have positive effects by enhancing effectiveness or negative effects by contributing to ADEs and toxicity (7). In addition, DDIs might reduce therapeutic efficacy by exerting inhibitory or inductive effects on cytochrome P450 (8). The risk of ADEs arises from DDIs when multiple medications are taken (9). Therefore, it is crucial to evaluate the possibility and severity of DDIs, as well as to identify the factors that influence them.

T2DM patients often have a high prevalence of chronic comorbidities and frequently experience polypharmacy (10, 11), putting them at a high risk for DDIs especially during hospitalization. Previous studies have retrospectively determined that the prevalence of pDDIs in T2DM patients varied from 10% to 81% in different studies (12–14). However, few studies have considered the influence of herbs or Chinese patent medicines, which are composed of several Chinese medicinal herbs with high consumption in patients with T2DM. Hence, we aimed to retrospectively investigate the epidemiology of pDDIs involving both chemical drugs and herbs or Chinese patent medicines, and to classify the factors influencing pDDIs (pCDIs+pHDIs) in T2DM hospitalized patients, both during hospitalization and after discharge.

A total of 6796 pDDIs (6674 pCDIs and 122 pHDIs) were identified during hospitalization, while 1753 (1727 and 26) were observed after discharge among T2DM inpatients. The prevalence of pDDIs was 99.73% during their hospital stay and 85.21% after discharge, which were significantly higher than those reported in previous similar studies (14, 15). A retrospective study conducted by Ivana Samardzic et al. revealed that 80.9% of patients with diabetes mellitus were exposed to at least one pDDI, but only category C interactions were detected by the Lexi-Interact software (14). While Ilona Ikäheimo et al. found that clinically relevant DDIs were presented in 44.5% of patients aged ≥65 years with T2DM in Finnish home-dwelling primary care (15). However, direct comparison between these studies was challenging due to discrepancies in study design, prescribed medications, pDDIs checkers used, criteria for identifying pDDIs and other factors (16). The majority of the identified pDDIs were associated with antidiabetic drugs both during hospitalization and after discharge. Approximately 31.90% of the total identified pDDIs during hospitalization involved interactions between hypoglycemic medications, while this percentage increased to 58.47% after discharge. The higher number of pDDIs observed among inpatients compared to those after discharge demonstrated an increased risk of DDIs during hospitalization. Nevertheless, more importantly, not all pDDIs would actually occur, as they are only considered as potential hazards. Additionally, some patients may encounter new diseases or complications after discharge from the hospital and may take additional medications that could initiate new pDDIs. Therefore, minimizing the risk of pDDIs is of the essence in T2DM patients.

Despite most pDDIs reported in our study had a C risk rating with moderate severity, more attention was warranted regarding the higher severity classifications of pDDIs. Among T2DM inpatients, the most frequently identified pCDIs in category X were antihistamines and potassium chloride, resulting in adverse reactions and an increased risk of ulcerative effects. Interactions with nonsteroidal anti-inflammatory drugs (NSAIDs) such as ibuprofen, celecoxib, indomethacin, and celecoxib were detected in our study population under category X. Some studies have suggested that regardless of the chemical class, all NSAIDs could produce augmented dose-dependent analgesia and gastromucosal injury with associated mucosal barrier dysfunction after extended use (2 weeks) (17). The concurrent use of NSAIDs, which represents therapeutic duplication (18), should be circumvented due to the increased toxic effects, especially gastrointestinal toxicity. Unfortunately, we have not collected any information on whether patients experienced gastrointestinal consequences as a result of this combination. The PD interaction between omeprazole and clopidogrel was the most frequently observed pDDI under category D among non-antidiabetic drugs. It was identified in 14 patients (1.54%) during hospitalization or in 7 patients (1.55%) after discharge, with the risk of therapy failure and a loss of clopidogrel’s protective cardiovascular benefits. Omeprazole could reduce the antithrombotic efficacy of clopidogrel by inhibiting the activity CYP2C19, resulting in lower serum concentrations of the active metabolite (19, 20). Studies in a multi-ethnic Asian population had shown that the concomitant use of omeprazole with clopidogrel was associated with an increased risk of myocardial infarction, but not with mortality or stroke (21). Thus, it was strongly recommended to replace omeprazole with rabeprazole or pantoprazole for patients using clopidogrel, as rabeprazole and pantoprazole had low CYP2C19 inhibitory potential (22, 23).

In our study, the proportion of T2DM inpatients who concurrently used Chinese patent drugs decreased from 7.46% during hospitalization to 2.44% after discharge. Among the pHDIs in X category retrieved from hospitalized patients, the combination of danshen and warfarin was found to be associated with an increased risk of bleeding effects (24). This was attributed to the enhanced bioavailability of both R- and S-warfarin when used concomitantly with danshen, leading to an exaggerated anticoagulant response (25). Previous reports had indicated that patients receiving warfarin therapy should avoid consuming herbal products, including danshen, ginseng, ginkgo, and dong quai because of their high risk of bleeding (26, 27). The major tanshinones in danshen had been reported to elevate the steady-state plasma concentration of warfarin by 23% through inhibition of warfarin hydroxylation (28, 29). Nonetheless, safe coadministration might be achievable with close monitoring and adjustment of the warfarin dosage. Specific herbs with antiplatelet activity, such as ginkgo, ginseng, and licorice, were detected under category D among non-antidiabetic drugs that could interact with clopidogrel or aspirin, potentially increasing the risk of bleeding. Although laboratory studies had indicated the potential for pHDIs between these herbs and aspirin or clopidogrel, there may not necessarily be a significant clinical correlation (30–32). However, until more evidence regarding their safety is available, it is advisable to avoid co-administration of these herbs with aspirin or clopidogrel.

This study showed a statistically significant correlation between the number of medications and the risk of pDDIs, both during hospital stays and after discharge. Previous studies had also found that the risk of DDIs increased with the number of medications, especially when they were from diverse pharmacological categories. This could be expected as numerous drugs had overlapping elimination pathways and non-selective mechanisms of action. Other factors associated with pDDIs after discharge included complications of T2DM. T2DM could induce various complications, resulting in an increased number of prescribed medications and consequently leading to an elevated risk of pDDIs.

However, there were some limitations in this study. First of all, only one pDDIs checker was utilized to evaluate the prevalence of pDDIs. Employing multiple pDDIs checkers could potentially improve the accuracy of identifying these interactions. Secondly, this study had a retrospective design and was conducted at a single center. It could not track the medications that patients self-administered after discharged or investigate the relationship between clinical outcomes and pDDIs. Although future multicenter and prospective studies in clinical settings were still needed to address all these limitations, this study was the first to elucidate the issues concerning pCDIs and pHDIs in T2DM patients, both during hospitalization and after discharge.

In this study, T2DM patients were exposed to a significant number of pDDIs, including pCDIs and pHDIs, both during hospitalization and after discharge. The concurrent use of multiple drugs was identified as the most significant risk factor for pDDIs. The utilization of drug interaction software, electronic warning systems, and prescription pre-review systems, along with the collaboration of clinical pharmacists, has the potential to dramatically reduce the potentially harmful drug combinations and contribute to enhancing patient safety. The development of universally applicable clinical guidelines for pDDIs may efficiently recognize pDDIs and provide further evidence to support clinically rational drug use.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving humans were approved by Jinshan Hospital of Fudan University Ethics Committee. The studies were conducted in accordance with the local legislation and institutional requirements. Written informed consent for participation was not required from the participants or the participants’ legal guardians/next of kin in accordance with the national legislation and institutional requirements. Written informed consent was obtained from the individual(s) for the publication of any potentially identifiable images or data included in this article.

WR: Resources, Data curation, Writing – original draft, Investigation, Funding acquisition. YL: Investigation, Formal analysis, Writing – review & editing, Software. HJ: Writing – review & editing, Investigation. XL: Formal analysis, Writing – review & editing, Supervision. NZ: Writing – review & editing, Supervision.