A comparative analysis of the drug interaction programmes amongst geriatric outpatients

INTRODUCTION

Geriatric populations exhibit a greater propensity for experiencing adverse drug reactions (ADRs) in comparison to other age groups, and diagnosing these adverse effects in older adults is challenging. The unusual symptoms that manifest further compound this complexity, ascertaining whether they are attributable to the ageing process or the medications employed for therapeutic purposes.^[1] The prevalence of ADRs in this population fails to evince any significant deviation from the previous times. The average rate of hospital admissions due to ADRs amongst the geriatric populations was recorded at 16.6%, in contrast to a significantly lower rate of 4.1% observed in younger individuals.^[2] The occurrence of drug-drug interactions (DDIs) is a major cause of ADRs. Older adults exhibit heightened vulnerability to the occurrence of DDIs due to several variables such as advanced age, multiple comorbidities and polypharmacotherapy.^[3]

The prevalence of possible DDIs amongst older adults with multimorbidity in primary care settings ranges from 20% to 100%.^[4] A systematic review has indicated that DDIs accounted for 0.57% of hospital admissions in the general population but for the geriatric cohort, this percentage increased to 4.8%.^[5] However, it is crucial to prioritise the screening of DDIs, as they are preventable and manageable adverse events.^[6]

Drug interaction programs are indispensable tools that healthcare professionals use to screen for potential risks that may arise from drug combinations. The vast amount of DDI information poses a challenge for clinicians to remember all of it. At present, there are several drug interaction programmes available in various forms, such as open access or subscription resources. Upon comparing the results of DDIs screening from different programmes, it has been observed that they may yield different outcomes.^[7] Former research has demonstrated a comparatively low level of consensus on the classification of DDIs across different applications.^[8,9] It is essential for the clinician to be aware of the varying results from each database and to consider their benefits and limitations when choosing to utilise them.^[10] Numerous investigations have been conducted to compare various drug interaction software applications, such as Lexicomp, Micromedex, Medscape, Eporactes, DDInter and Drugs. com.^[11-13] The software applications most commonly selected by clinicians in hospital settings are Lexicomp, Micromedex and Medscape. The majority of the studies found that Lexicomp received the highest scores. However, it was found that previous studies were conducted in oral oncolytic ^[12,14] organ transplants ^[7,13] and antiretroviral drug interactions.^[15] However, a comparative analysis of all three programmes in the elderly population remains limited. Therefore, the aim of this study was to compare drug interaction programmes in the geriatric population at an outpatient clinic of a tertiary care hospital, with the objective of assisting physicians and pharmacists in selecting the most appropriate programme for screening purposes to ensure accuracy and safety in medication prescribing for older patients.

MATERIALS AND METHODS

RESULTS

Prevalence of DDIs using Lexicomp, Micromedex and Medscape

A cumulative cohort of 10,877 individuals was gathered for participation in this research through electronic medical records. The mean age of patients was 74.3 (standard deviation 6.8 years), characterised by a notable prevalence of male subjects (54.3%). Table 3 illustrates the overall prevalence of DDIs in the geriatric population, with a focus on moderate and major DDIs, as well as contraindications, as analysed by Lexicomp, Micromedex and Medscape. In this study, it was observed that drugs that were deemed contraindicated were not included in any of the three-drug interaction programmes. Our findings indicated that while the overall figures were similar, differences in severity were observed. Figure 1 further represents the overlap in the prevalence of major DDIs, as identified by the Micromedex, Medscape and Lexicomp Drug Interaction Programmes. There were only 1700 major DDIs (15.6%) that exhibited consistency across all three drug interaction programmes. The majority of major DDIs were found in Micromedex, followed by Lexicomp and Medscape. The prevalence of major DDIs in Micromedex was not correlated with other programmes, resulting in a total of 3202 DDIs or 29.4%. The major DDIs between Lexicomp versus Micromedex exhibited consistency, with 1156 DDIs or 10.6%. Similarly, Micromedex versus Medscape accounted for 243 DDIs or 2.2%, and Lexicomp versus Medscape was found to be 9 DDIs or 0.1%.

Table 3: Prevalence of potential drug–drug interactions amongst drug interaction programmes.

Category	Lexicomp	Micromedex	Medscape
Minor+None, n (%)	3858 (35.5)	3495 (32.1)	3355 (30.8)
Moderate, n (%)	3960 (36.4)	1081 (10.0)	5543 (51.0)
Major, n (%)	3059 (28.1)	6301 (57.9)	1979 (18.2)

Close

Figure 1:

Prevalence of major drug–drug interaction identified by the Micromedex, Medscape and Lexicomp Drug Interaction Programmes.

Export to PPT

The three most encountered DDIs with major severity were omeprazole versus clopidogrel (6.3%), followed by central nervous system (CNS) depressants versus opioid analgesics (4.2%), and orphenadrine versus CNS depressant (3.2%). Various programmes detected different severity classifications. For example, orphenadrine versus CNS depressant was detected with major severity in Lexicomp but none in Micromedex and was classified as moderate in Medscape and prednisolone versus aspirin was detected with major severity in Micromedex but was deemed moderate in Lexicomp and Medscape [Table 4].

Table 4: Interacting pair amongst drug interaction programme and severity.

Interacting pair		n (%)	Interaction effect	Severity
				Lexicomp	Micromedex	Medscape
Omeprazole	Clopidogrel	682 (6.3)	Reduce serum concentrations of the active metabolite (s) of clopidogrel	Major	Major	Major
CNS depressant	Opioid analgesics	460 (4.2)	Augment the CNS depressant effect of opioid agonist	Major	Major	Moderate
Orphenadrine	CNS depressant	349 (3.2)	Augment the CNS depressant effect of orphenadrine	Major	None	Moderate
TCA	NSAIDs	199 (1.8)	Result risk of bleeding	Moderate	Major	None
Methotrexate	NSAIDs	198 (1.8)	Increase the serum concentration of methotrexate	Major	Major	Major
Prednisolone	Aspirin	136 (1.25)	Salicylates may augment the adverse/ toxic effect of prednisolone.	Moderate	Major	Moderate
Omeprazole	Rifampicin	126 (1.16)	CYP2C19 inducers may reduce the serum concentration of omeprazole	Major	Major	Major

CNS: Central nervous system, TCA: Trans-cinnamaldehyde, NSAIDs: Non-steroidal anti-inflammatory drugs

DISCUSSION

The prevalence of DDIs in older patients varies depending on the study setting, utilised programmes and selected population. This study involved analysing the medication record in older patients using three-drug interaction programmes at an outpatient facility of a tertiary care hospital. The prevalence of major DDIs in older patients using Lexicomp, Micromedex and Medscape in this study was 28.1%, 57.9% and 18.2%, respectively. Our findings were different from one report in China, and it showed that the prevalence of major DDIs in comparable settings based on three databases such as Lexicomp, Micromedex, and DDInter, was 32.2%, 32.9% and 22.6%, respectively.^[4] A study in the nephrology outpatient setting of a University hospital in Hong Kong reported that the prevalence of major and severe DDIs using Lexicomp, Micromedex, and Medscape was 19.6%, 37.8% and 35.5%.^[13] These findings indicated weak compatibility among the databases, as evidenced by the interrater agreement obtained from the programmes in our study, particularly with regard to major DDIs (which showed only slight agreement). The possible reasons for the disparity amongst drug interaction programmes such as Micromedex, Medscape and Lexicomp might be due to the variations in the number of potential DDIs detected, the severity of these interactions and the overall quality of information provided. These differences could be attributed to database algorithms, update frequency and the underlying sources of drug interaction data.^[23-25] Nonetheless, our study demonstrated substantial agreement between the minor DDIs and the no-interaction group. It implies a high negative predictive value of the DDI report while a low sensitivity in detecting major DDIs. Therefore, it is essential for healthcare practitioners to possess a comprehensive understanding of the diverse potential DDIs identified by various drug interaction software programs.

Drug interaction programmes utilise scientific and contemporary literature to compile information into categories that are accessible to healthcare professionals. However, the results from this study revealed the discrepancies of major DDIs across the three databases, and in actual clinical practice, the individual patient has different background characteristics that might not have been covered in the previous studies. In addition, the diverse factors of each patient cannot be fully integrated into these programmes. Therefore, the use of more than one database in detecting DDIs, particularly for major DDIs, is suggested to ensure a thorough assessment and identification of potential drug interactions. Furthermore, the establishment of clinical protocols concerning the occurrence of clinically substantial DDIs and their probable unfavourable consequences, along with the provision of management approaches to aid clinicians in the prompt identification of potential DDIs, necessitates a thorough evaluation of relevant literature and consultations with proficient clinical pharmacists and experienced physicians. The successful execution of such initiatives can avert the oversight of these interactions and promote the appropriate administration of pharmacotherapy for geriatric patients.^[4,26,27]

The frequent major DDIs of the three-drug interaction programmes in this research were omeprazole versus clopidogrel, followed by CNS depressants versus opioid analgesics and orphenadrine versus CNS depressant. These findings should alert physicians and clinical pharmacists, especially while prescribing these medications. Studies have been carried out in Thailand that have produced similar results.^[28] Despite pharmacodynamic studies indicating that omeprazole possesses the ability to alleviate the antiplatelet effect of clopidogrel. Nevertheless, the evidence from observational studies does not demonstrate an increase in cardiovascular risk among patients undergoing treatment with this combination. ^[29,30] Patients treated with the combination of clopidogrel and omeprazole displayed a reduction in the likelihood of experiencing gastrointestinal events and a similar likelihood of experiencing cardiovascular events. However, the judgement to administer proton-pump inhibitors to patients receiving clopidogrel should be made on an individualised basis, taking into account the patient’s bleeding and cardiovascular risk factors.^[31] In the United States, studies have indicated that the combination of CNS depressants and opioid analgesics, or the co-administration of two CNS depressants (including muscle relaxants), was the most commonly found DDI in the elderly population.^[32] It is advisable to refrain from employing simultaneous administration of agents that depress the CNS unless there are no feasible alternative therapies. In the event that concomitant use is necessary, the healthcare professional should administer the minimal feasible amount and period of each medication, all the while guaranteeing that the intended therapeutic outcome is achieved.^[33]

The utilisation of a dependable and extensively validated database derived from preceding research was a key strength of this investigation.^[4,34] In addition, it was the first study to compare the prevalence of DDIs in a geriatric outpatient setting within a tertiary care hospital in Thailand. However, some limitations of this study were identified. First, because of its retrospective nature, the data acquired from electronic medical records may be deficient or may underrepresent certain aspects, particularly the data concerning the actual clinical complication of major DDI. In addition, data concerning the usage of over-the-counter medications or herbals were not gathered. Second, this study was performed within a single institution, which might limit the generalisability in other settings. Third, the database has undertaken multiple revises concerning potential DDIs. Therefore, if any changes have been enacted with respect to particular pharmaceuticals across each category, the outcomes of this study may likely differ from its prior versions.

CONCLUSION

The level of agreement amongst the major DDI programmes, namely Lexicomp, Micromedex and Medscape, was slight. To promote consistency in drug information references, it is imperative that standardisation measures are put in place for DDI documentation. It is essential for healthcare providers to acknowledge this variability and understand that these resources only contribute partially to the overall quality of prescribing. To facilitate the timely detection of potential DDIs, it is recommended that a clinical practice guideline should be developed to address clinically significant DDIs and provide management strategies for clinicians.