Antibiotic-Related Adverse Drug Reactions at a Tertiary Care Hospital in South Korea

Background Adverse drug reactions (ADRs) are any unwanted/uncomfortable effects from medication resulting in physical, mental, and functional injuries. Antibiotics account for up to 40.9% of ADRs and are associated with several serious outcomes. However, few reports on ADRs have evaluated only antimicrobial agents. In this study, we investigated antibiotic-related ADRs at a tertiary care hospital in South Korea. Methods This is a retrospective cohort study that evaluated ADRs to antibiotics that were reported at a 2400-bed tertiary care hospital in 2015. ADRs reported by physicians, pharmacists, and nurses were reviewed. Clinical information reported ADRs, type of antibiotic, causality assessment, and complications were evaluated. Results 1,277 (62.8%) patients were considered antibiotic-related ADRs based on the World Health Organization-Uppsala Monitoring Center criteria (certain, 2.2%; probable, 35.7%; and possible, 62.1%). Totally, 44 (3.4%) patients experienced serious ADRs. Penicillin and quinolones were the most common drugs reported to induce ADRs (both 16.0%), followed by third-generation cephalosporins (14.9%). The most frequently experienced side effects were skin manifestations (45.1%) followed by gastrointestinal disorders (32.6%). Conclusion Penicillin and quinolones are the most common causative antibiotics for ADRs and skin manifestations were the most frequently experienced symptom.


Introduction
Adverse drug reactions (ADRs) are any unwanted/uncomfortable effects from medication resulting in physical, mental, and functional injuries [1]. ADRs experienced by hospitalized patients are associated with increased morbidity and mortality, prolonged hospitalization, and increased medical expense [2]. For this reason, several studies have suggested that ADRs are a major public health concern [3].
Disease prevalence, economic status, culture, and ethnicity all contribute to different ADR patterns [4,5]. The overall incidence of ADRs varies by study but ranges from 0.15% to 30% [1,6]. In one study conducted at an Indian tertiary care hospital, antibiotics were responsible for 40.9% of ADRs [6]. An Australian tertiary center reported that antibiotics were related to 25% of ADRs [7]. Furthermore, previous studies have shown that 26.88% of ADRs are considered severe and that 99.47% required additional medical intervention [8].
Many observational studies have examined the incidence, pattern, and severity of ADRs, but most of these have been performed in America or Europe; reports on Asian countries are extremely rare [7,9,10]. Several South Korean reports have identified antibiotics as a leading cause of ADRs, but most are based on information from primary care center pharmacies, and data on ADRs related to antimicrobial agents reported from tertiary care hospitals are extremely rare.
Although a number of studies on ADRs caused by various drugs have been conducted, none have focused specifically on antibiotics. Therefore, in this study we investigated the 2 BioMed Research International frequency of antibiotic-related ADRs experienced at a tertiary health care hospital in South Korea.

Study Design.
This was a retrospective cohort study based on reports from Yonsei University College of Medicine Severance Hospital, a tertiary health care hospital in Seoul, South Korea, from January 1 to December 31, 2015. Only antibiotic-related ADRs in hospitalized patients were analyzed. All antibiotics, whether administered concurrently or at a different time point, were evaluated for possibilities of ADRs and included in the analysis. Any cases of ADRs that might have been caused by concurrently administered drugs, other than antibiotics, were excluded from the analysis.
The following data were collected: date of reported ADR, age, gender, clinical manifestation, causal drug and brand name, route of administration, dates of administration and discontinuation, outcome (serious or not serious), recurrence, causality assessment, and dose-relationship. This study was approved by the Institutional Review Board (IRB) of Severance Hospital (IRB #4-2017-0307), and the need for written informed consent from all participants was waived by the approving IRB.

Definitions.
Causality was classified into three categories: certain, probable, and possible based on the WHO-Uppsala Monitoring Center criteria [11,12]. The severity of each ADR was classified as serious or nonserious [12]. Serious ADRs were defined as patients who experienced disability, prolonged hospitalization, life-threatening symptoms, or death [12]. Symptoms were classified according to symptom organ class (SOC) from the Medical Dictionary for Regulatory Activities (MedDRA) [13]. Defined daily dose (DDD) is the average maintenance dose per day for a drug used for its main purpose, as defined by the World Health Organization (WHO) [14]. Antimicrobial use density (AUD) describes the total antimicrobial use in DDD per 1,000 patient days of one drug class, as recommended by the WHO [14]. AUD was calculated as follows. AUD = (total antimicrobial use)/(DDD × patient days) × 1,000 [15,16].

Collected Data and Reporting
Sources. Severance Hospital is a 2400-bed tertiary care hospital and is one of the largest health care centers in South Korea. Severance was registered as a Regional Pharmacovigilance Center in 2006 and is using a computer-based pharmacovigilance monitoring system. ADR reporting is voluntary and can be reported by a physician, pharmacist, nurse, or patient who recognizes the ADR event. These voluntary reports are reviewed by the ADR-monitoring team, which includes a physician from the Department of Allergy and Clinical Immunology and a pharmacist. Then the clinical and demographic information of the reported ADR is stored in a pharmacovigilance system database and noted in the patient's electronic medical record (EMR). The computerized system improves medication safety by alerting medical practitioners to drug allergies and any drug-drug interactions the patient experienced.

Data Analysis.
Descriptive statistic procedures were performed to analyze the ADR cases. Categorical variables are presented as numbers and percentages. All statistical tests were performed using SPSS 18.0 (Statistical Package for the Social Sciences, Chicago, IL, USA).

Demographic Data, Severity, and Causality.
In total, 2,032 cases of antibiotic-related ADRs were reported during the study period. Of these, 1,277 (62.8%) were proven to be antibiotic-related based on the World Health Organization-(WHO-) Uppsala Monitoring Center criteria. The median age was 54 years (range 35-78), and 610 (47.8%) patients were male. Causality assessment based on WHO criteria revealed that 28 (2.2%) cases were certainly caused by antibiotics, 456 (35.7%) were probably caused by them, and 793 (62.1%) were possibly caused by them ( Figure 1). A severity assessment confirmed 44 (3.4%) serious ADRs. Death or life-threatening events, hospital admission or prolonged hospital stay, or disability occurred in 2 cases (4.5%), 38 cases (86.3%), and 4 cases (9.0%), respectively.

Frequency of ADRs by Symptom and the Most Common
Causative Antibiotics. Skin and subcutaneous tissue disorders were the most common clinical manifestation, occurring in 576 cases (45.1%), followed by gastrointestinal disorders, which occurred in 416 cases (32.6%) ( Figure 2).

Antimicrobial Use Density (AUD) to Demonstrate Each
Class of Antibiotics Usage. In our study, the antibiotic uses of penicillin were 2,179.2 AUDs, followed by third-generation cephalosporin and quinolone with AUDs of 1,277.8 and 837.9, respectively (Supplementary Table 1).

Discussion
Several South Korean reports have identified antibiotics as the most common cause of ADRs [17,18]. However, most of these reports have been based on data from private clinics and pharmacies rather than tertiary care hospitals. Here, we report the antibiotic-related ADRs experienced at a tertiary care hospital. In this study, 3.4% of patients experienced serious ADRs. One multicenter study conducted in 2009 covering six Regional Pharmacovigilance Centers in South Korea reported that 17.7% of ADRs were serious [18]. A metaanalysis reported that 6.7% of ADRs were serious and that 0.32% of ADRs were fatal [19]. However, it is difficult to compare these results with our study because the previous studies included nonantibiotics such as nonsteroidal antiinflammatory drugs (NSAIDs) and radiocontrast media.
Antibiotics have been reported to be major causes of ADRs [20]. In a study that only included outpatients, sulfonamides followed by penicillin were reported to be the most common causative antibiotics [20]. Prior reports have shown that quinolones, ciprofloxacin in particular, are another common causative antibiotic [21]. This study shows that penicillin and quinolones were responsible for the majority of ADRs. These results are similar to several other South Korean reports [18,22].
Geer et al. [6] reported that antituberculosis drugs accounted for 13.15% of all ADRs, and Maciel et al. [23] reported that up to 83.54% of ADRs were caused by antituberculosis drugs. In a study in Iran, gastrointestinal symptoms (22%) and hepatotoxicity (35.7%) were frequently    experienced ADRs caused by antituberculosis drugs [24]. In this study, antituberculosis medications made up a smaller proportion (61 cases 4.8%) of ADRs; however, gastrointestinal reactions (11.5%) and hepatotoxicity (9.8%) were both common symptoms experienced in our study, which is similar to the results of previous studies. Isoniazid was accountable for nausea/vomiting in 2 cases, hepatobiliary disorders in 4 cases, and skin and subcutaneous tissue disorders in 8 cases, and 1 case was associated with anaphylaxis. Rifampin was accountable for nausea/vomiting, skin and subcutaneous tissue disorders, and allergic disorders in 3, 9, and 4 cases, respectively. 1 case was associated with rifampin induced Stevens-Johnson syndrome. Ethambutol ADRs were mostly associated with skin and subcutaneous tissue disorders (13 cases), and ethambutol induced optic neuritis was confirmed in 4 cases. The majority of pyrazinamide ADRs were also skin and subcutaneous tissue disorders, 5 cases.
Of all cutaneous ADRs considered in a previous study, antibiotics were the main cause (46.55%) [25]; in another study, antibiotics accounted for 48% of delayed cutaneous ADRs, 20% of which were purportedly due to glycopeptides and sulfonamides [26]. In particular, glycopeptides and sulfonamides were implicated in 20% of these ADRs [26]. In our study, 45.1% of skin and soft tissue ADRs were due to antimicrobial agents. Quinolones, penicillin, thirdgeneration cephalosporins, and glycopeptides were the most common causative antibiotics for skin and subcutaneousrelated ADRs. The difference in causative antibiotics may be explained by the ethnicities included in each study [5]. Further studies on the mechanisms behind causative antibiotics and reactions are needed.
Penicillin allergies are more common in females [27], as is the frequency of ADRs [28]. We also found a slight female predominance in our study (47.8% of patients who experienced ADRs were male).
There were several limitations to our study. First, it was a single-center study and lacked reports from private clinics and other Asian countries. Further studies regarding antibiotics and ADRs are necessary to validate our results and provide more generalizable data covering all Asian countries. Second, reports of ADRs are voluntary at our hospital, so many cases could have gone unreported. Third, only data on hospitalized patients were collected; ADRs of outpatients were not included in the study. Finally, DDD is a unit of measurement and does not necessarily reflect the recommended dose or prescribed daily dose (PDD). The PDD for each class of antibiotic was not reported by the pharmacovigilance monitoring system used in this study. As there is a known discrepancy between the PDD and the daily DDD, further validation by PDDs would be necessary for accurate comparisons between antibiotics.

Conclusions
In conclusion, penicillin and quinolones were the most common antibiotic causes of ADRs. The most frequently experienced clinical feature was skin manifestations. These findings may help identify patterns and causative antibiotics of ADRs in Asian countries.

Conflicts of Interest
The authors declare they have no conflicts of interest.

Authors' Contributions
Nam Su Ku and Jung Won Park contributed equally to this article.