Assessment of potential drug-drug interactions in hospitalized patients with infectious diseases: an experience from a secondary care hospital

Introduction

Infectious infections are among the most common health concerns globally, regardless of age. Infected people frequently require hospitalization, which increases the risk of morbidity and mortality and raises healthcare costs. Infectious diseases are conditions caused by microscopic organisms such as bacteria, viruses, fungi, or parasites that spread from one person to another.¹ Indeed, healthcare providers frequently face challenges in selecting and using antimicrobial medicines.^2,3

Drug-drug interactions (DDIs) occur when two or more co-administered drugs interact, with one drug altering the effect of the co-administered drug. The outcome effect of drug interactions may vary from non-serious to serious/life-threatening (or) irreversible, affecting the goals of therapy, clinical effectiveness, and worsening treatment outcomes.⁴ Studies have reported that age (≥ 65 years), polypharmacy, increased number of prescribers, and comorbid illness are defined risk factors for drug interactions. In addition, a decline in drug metabolism associated with aging, comorbidities such as hepatic and renal injury, and altered drug plasma concentrations complicate the medication use process and increase the sensitivity to drug interactions. As a result, clinically significant drug interactions prolong hospital stays, increase re-visit and healthcare expenditures, and aggravate patient outcomes in inpatient and outpatient healthcare settings.^5–7

Drug interactions are classified as pharmacokinetic and pharmacodynamic interactions; a few are categorized as unknown or other based on their mechanism of interactions. Drug interactions can be grouped as major, moderate, or minor according to severity and significance.⁸ Studies carried out in different health settings and patients reported that potential drug-drug interactions (pDDIs) range from 19.3% to 91.6%.^9,10 A systematic review and meta-analysis reported that the prevalence of clinically manifested DDIs ranged from 1.2% to 64.0%.¹¹ The increased incidence of adverse outcomes associated with drug-drug interactions is a common cause of hospital admission, primarily in the aging population.¹² However, the variation in the results across different studies is associated with factors such as patient characteristics, prescribing pattern, severity of the illness, study population, and study setting.

The use of clinical decision support systems, close monitoring of patient's drug therapy, and involvement of clinical pharmacists in a multidisciplinary team are some of the important measures that help to minimize drug interactions and improve patient safety.^13,14

Studies on antimicrobial agents in the United Arab Emirates (UAE) have focused on the prescription pattern of drug use and related outcomes in various hospital settings. However, studies related to DDIs with antimicrobial agents in infectious diseases are unaddressed despite being one of the reasons for hospitalization. Therefore, the present study was carried out to assess pDDIs among hospitalized patients with infectious diseases in a secondary care hospital.

Methods

Results

Class of medications involved in causing drug interactions

The analysis of the class of medications involved in the onset of pDDIs carried out using the Anatomical Therapeutic Chemical (ATC) classification system derived by the World Health Organization found a higher prevalence in the category cardiovascular system (28.8%) followed by anti-infective for systemic use (23.9%) and alimentary tract and metabolism (21.5%) (Figure 1).

Figure 1. Anatomical Therapeutic Chemical (ATC) classification of drugs involved in potential drug-drug interactions (pDDIs).

The evaluation of the underlying mechanism that causes DDIs showed that 210 (38.25) interactions involve pharmacokinetic interactions, while 181 (32.96%) interactions were caused by ‘others’ or unknown mechanisms. The remaining 158 (28.77%) interactions were known to be produced by pharmacodynamic interactions.

The most frequently identified DDIs were the combination of atorvastatin with clopidogrel, bisoprolol, amlodipine, or pantoprazole; aspirin with insulin; clopidogrel enoxaparin and furosemide with valsartan; and lisinopril and bisoprolol. The antimicrobial drugs involved in pDDIs were combinations of ceftriaxone with enoxaparin and aspirin, levofloxacin with insulin, furosemide, and piperacillin/tazobactam with aspirin, metformin, and doxycycline (Table 4).

Table 4. Most common drug-drug interactions and their effect were identified among the study population.

Interaction	Drug combinations	Frequency	Effect	Level of concern	Source of recommendation
Significant	Aspirin – Enoxaparin	04	Both increase anticoagulation & Increased risk of bleeding	04	pa
	Clopidogrel – Enoxaparin	08	Both increase anticoagulation & Increased risk of bleeding	04	pa
	Bisoprolol – potassium chloride	04	Both increase serum potassium	04	a
	Valsartan – Sacubitril	04	Increased risk of renal impairment, hyperkalemia, hypotension,	04	a
	Metronidazole – levofloxacin	05	METRONIDAZOLE and LEVOFLOXACIN both increase QTc interval, Increased risk of long QT Syndrome, and possible Torsades de pointes	04	p
Moderate	Isoniazid – Rifampicin	05	Rifampin enhances the metabolism of isoniazid to hepatotoxic metabolites	03	a
	Levofloxacin – Furosemide	06	Increased risk of long QT Syndrome and possible Torsades de pointes	03	pa
	Bisoprolol – Aspirin	13	Both increase serum potassium	03	pa
	Amlodipine – Atorvastatin	16	Both levels probably increased – increased risk of arrhythmia, edema, myopathy, elevated liver function tests	03	pa
	Levodopa – piperacillin/tazobactam	04	Anticholinergics may enhance the therapeutic effects of LEVODOPA but may also exacerbate tardive dyskinesia. In high doses, anticholinergics may decrease the impact of LEVODOPA by delaying its GI absorption	03	a
Minor	Clopidogrel – Atorvastatin	14	Levels of Clopidogrel's active metabolite can be decreased	02	pa
	Insulin – levofloxacin	08	Insulin effects may be increased, and Quinolone antibiotic administration may result in hyper- or hypoglycemia	02	a
	Lisinopril – levodopa	03	LISINOPRIL effects may be increased Consider decreasing the dosage of an antihypertensive agent	02	a
	Aspirin – piperacillin/tazobactam	16	Both Piperacillin/Tazobactam (PIPERACILLIN) and ASPIRIN levels may be increased	02	a
	Clopidogrel – Amlodipine	05	Levels of CLOPIDOGREL's active metabolite can be decreased	02	pa
Minor/non-significant	Bisoprolol/atorvastatin	17	Bisoprolol levels can be slightly increased, and Increased risk of bradycardia	01	pa
	Furosemide/piperacillin & tazobactam	06	Both decrease cholinergic effects/transmission Increased risk of anticholinergic syndrome (dilated pupils, vasodilation/flushing, hyperthermia, dry skin)	01	p
	Metformin – furosemide	02	METFORMIN levels may be increased	01	a
	Memantine – metformin	01	Both drugs minimally increase the effects of the other drug involved in the mechanism	01	pa
	Clindamycin – piperacillin/tazobactam	02	CLINDAMYCIN and Piperacillin/Tazobactam (PIPERACILLIN) both decrease cholinergic effects/transmission Increased risk of anticholinergic syndrome (dilated pupils, vasodilation/flushing, hyperthermia, dry skin, hallucinations/agitation, constipation/urinary retention, tachycardia)	01	p

Discussion

Drug interactions contribute to undesirable health outcomes, compromise the clinical effectiveness of drug therapy, increase hospital visits, and prolong hospital stays.¹⁵ The overall prevalence of pDDIs in our study was 67%, higher compared to the study by Hamdouk et al., who reported at least one pDDIs in 62.9%8% of the study sample.¹⁶ Downward trends in prevalence were documented in earlier studies by Kuscu et al. (60%) and Rabba et al. (56%), respectively.^17,18

This disparity in the prevalence of pDDIs may be attributed to the differences in the study setting, study population, prescribing pattern of medications, and types of pDDIs and tools used to screen drug interactions in the study. In the present study, the average was 3.70±4.58 drug interactions per prescription among hospitalized patients. Documented evidence indicates that drug interactions occur more predominantly in hospitalized patients than in outpatients, considering the severity of the disease, comorbidities, and prescription of multiple medications with frequent modifications during their stay.¹⁹

The current study prescribed aspirin, clopidogrel, statins, enoxaparin, furosemide, valsartan, and bisoprolol to prevent and manage cardiovascular diseases. Documented evidence has reported that prescribing these medications alone or in combination is responsible for various interactions, such as increased bleeding, electrolyte imbalance, renal failure, and hypotension.^17,19–21 However, prescribing these medications is sometimes unavoidable and therapeutically valuable as a lifesaving medication. Therefore, close monitoring for effective treatment and evaluation of the benefit-risk assessment of actual DDIs of prescribed drugs is warranted. At the same time, careful laboratory assessment of international normalized ratio, serum electrolytes, renal and liver function tests, signs and symptoms of bleeding, and blood pressure monitoring are vital during treatment.

Similarly, metformin, sitagliptin, insulin, tamsulosin, memantine, levodopa, pantoprazole, paracetamol, and supplements such as potassium chloride and calcium carbonate were some of the important medications prescribed for the various other medical conditions in our study. In addition, drugs such as penicillins, cephalosporins, fluoroquinolones, metronidazole, macrolides, doxycycline, linezolid, isoniazid, rifampicin, vancomycin, and amphotericin B were some of the important antimicrobial agents used in this study. Drugs that cause enzyme induction or inhibition, resulting in reduced metabolism or clinical effects and alteration of gastrointestinal absorption are the most common mechanisms related to antimicrobial interactions.²²

A review of the published evidence suggests that fluoroquinolones and macrolides are associated with QT prolongation and may cause life-threatening arrhythmias.^17,23,24 Cautious prescribing should be exercised when co-administering drugs with a narrow therapeutic index and drugs metabolized through cytochrome P450 isozymes that can develop clinically significant unpredictable drug interactions, particularly in patients with renal and hepatic impairment and the elderly population. In the present study, the class of medications involved in pDDIs by the ATC classification showed a higher prevalence in the cardiovascular system (28.2%) followed by anti-infective for systemic use (23.9%). The increase in the prevalence of cardiovascular disease could be related to the use of complex medications for the long-term treatment of comorbidities and associated complications among the study populations. Our findings are consistent with those of Noor et al., Vazquez-Cornejo et al., and Samardzic et al., who reported an increased prevalence of pDDIs in patients with cardiovascular disease.^19,20,24 Furthermore, earlier study by Pavanello et al., in critical care patients showed that the most common drug class involved in pDDIs was anti-infective for systemic use, accounting for 45.8%, respectively.²⁵ The difference in study settings, varying profiles of study populations, disagreement in treatment guidelines and prescribing practice, and the use of different clinical decision support tools to analyze drug interactions may help explain the difference in the class of drugs involved in the onset of pDDIs.

In the present study, the severity level of most drug interactions was ‘moderate’ (40.0%). All pDDIs. are not identical in severity and evaluating their severity l is very important to recognize their clinical significance and proper management. Only small percentage (5.28%) of the identified drug interactions were found to be categorized as having a level of severity ‘significant’, which requires close monitoring to avoid any adverse outcome of the pDDIs.. A possible reason could be that physicians are aware of the risk factors and severity of pDDIs.and might have customized drug therapy individually to avoid or minimize pDDIs.

These findings align with the results of the study by Noor et al. and Obeid et al., who reported that most of the interactions were ‘moderate’ in severity.^20,26 Contrary to our findings, studies by Rabba et al. and Eneh et al. reported 66.4% and 52.7% of interactions with ‘major’ in severity level.^18,23 The difference in defining the classification and grading of severity between the resources could be a possible reason for the varying study results. Studies have observed that mechanism of action plays a significant role in DDIs, which requires management by either reducing the dose of one drug by 25% or 50%, changing the frequency of administration and dosage form, or avoiding such combination, replacing it with another medication.^19,27,28

Our study showed that ‘pharmacokinetic interactions’ were the most common underlying mechanism that caused pDDIs compared to pharmacodynamic and other unknown mechanisms. Similar observations have been cited in the study by Tesfaye et al., who reported pharmacokinetic interactions as the most common mechanism involved in causing pDDIs compared with pharmacodynamic and other/unknown interactions.²⁹

Studies have emphasized that patient characteristics such as age, comorbidities, number of medications prescribed, and hospital stay are risk factors for clinically significant pDDIs.^28,30,31 These findings align with our study observations. However, multiple binary logistic regression has revealed that comorbidities and number of drugs prescribed were the predictors associated with pDDIs among the study populations. It is important to note that aging populations are at risk of developing multiple comorbid medical conditions that require frequent hospital visits and a prolonged stay prescribed with more complex therapeutic regimens. Physiological changes related to age and variations in pharmacokinetics and pharmacodynamic parameters increase the risk and greater chance of developing pDDIs and adverse outcomes that reduce the efficacy of the treatment.³²

Conclusions

The present study identified a higher frequency of pDDIs in hospitalized patients with infectious diseases. Antimicrobial agents and co-prescribed medications interacted; most of the interactions in our study had ‘moderate’ levels of severity. This study highlighted that advanced age, multiple comorbidities, and polypharmacy were independent risk factors for pDDIs. Knowledge about pDDIs and the regular use of professional drug information database support systems can help prescribers optimize drug therapy and enhance health outcomes. The study strongly recommends that regular review of patient drug therapy by a clinical pharmacist might avoid possible drug combinations that are likely to cause pDDIs and could ring a bell in improving the quality of patient-centered care.