Multidrug interactions: the current clinical and pharmacovigilance challenge
Elizabeth Ellen Roughead · Journal of Pharmacy Practice and Research · 2015
New challenges for realising medication safety in practice, and for achieving effective pharmacovigilance, are arising because of the increasing prevalence of people living with multiple chronic health conditions. Forty percent of Australians now report living with three or more chronic illnesses, the majority of whom will be taking 5 or more medicines, with up to 20% taking 10 or more. The rising prevalence of polypharmacy has significant implications for medication safety, particularly in the area of drug interactions. A recent US study found that the number of drug-interaction alerts rose from 3 per person for those taking 5 medications, to 14 per person for people using 10 medicines.1 This issue is further complicated by new understandings in pharmacology. Historically, medicines were considered to interact with a single target and side effects not related to that mechanism were considered idiosyncratic. Studies using network analyses are enabling us to identify the interactions amongst medicines and proteins and their related molecular pathways across the whole body system. This branch of research, known as systems pharmacology, recognises that side effects of medicines may be related to the impact of medicines on multiple targets or a single drug target impacting on multiple cellular functions.2 The understandings from systems pharmacology will have implications for practice by improving the ability to predict multidrug interactions. For example, a patient with glaucoma, diabetes, hypertension and arthritis, may be prescribed a prostaglandin for glaucoma, a dipeptidyl peptidase inhibitor for diabetes, an ACE-inhibitor for hypertension, and an NSAID for arthritis, all of which have the potential to interact through their actions on different targets within the bradykinin–prostaglandin system. Drug-induced arrhythmia is another example where systems pharmacology approaches are being used to determine how and where the different medicines that affect the QT interval act within the system.3 One of the particular challenges in detecting and preventing multidrug interactions is the incremental nature of the effect with each additional therapy. This may be happening across a long time horizon in an aged population and so not considered collectively as a potential medication-related problem. For example, when a medicine with anticholinergic properties is started and a year later a second or third medicine with anticholinergic properties is added for an unrelated condition, the combined effect may not be considered as a possible cause of patient deterioration for an outcome such as confusion; yet the risk of hospital admission for confusion doubles for those using two medicines with anticholinergic properties and increases almost fourfold when treatment includes three or more medicines with anticholinergic activity.4 A similar issue arises with sedative medicines and risk of falls, where use of three sedative medicines concurrently doubles the risk of hospitalisation for falls.5 A focus on preventing potential interactions, even where the patient does not appear to be harmed by the interaction, is likely to assist in reducing future multidrug interactions. A particular example is the co-administration of medicines that are substrates and inducers of the same cytochrome P450 (CYP450) enzyme. A US study found that one-fifth of older people taking multiple medicines were co-prescribed medicines that were specific CYP450 inhibitors or inducers with substrates of the same enzyme. Further, 4% of people were on multiple medicines which were inhibitors of the same enzyme, and 30% of people using 10 or more medicines were treated with multiple medicines that interact across the same pathway.1 While pairs of medicines interacting across these pathways may be manageable by dose adjustment, each additional medicine adds to future complexity of dose and side effect management. It is concerning that alternative approaches using therapies that do not interact across the same pathway are not used, even when they are available. Australian data on the use of statins or proton pump inhibitors for people who had also been dispensed warfarin found that the most commonly co-prescribed statins or proton pump inhibitors were those with the potential to interact via the same CYP450 pathway, while the available medicines metabolised by different pathways were less commonly dispensed,6 suggesting a lack of awareness of this issue. The rising prevalence of people living with multiple chronic conditions requires a shift in focus from single drug side effects and drug–drug or drug–disease interactions, to multidrug interactions and the consequent effects. Guidelines will not be able to address all multidrug interactions that require consideration because of the unique medication profiles of patients on multiple medicines. Analysis of US data found that people 65 years and older taking three or more medicines had at least 98 unique medication regimens per 100 persons.1 Thus, active clinical pharmacist involvement in detecting and preventing multidrug interactions will be essential for improving care in patients with multiple chronic illnesses.