In a recent study published in PLOS Digital Health, researchers investigated the complexity of polypharmacy and whether frequent dose adjustments were required for co-medications and drug pairs in hospital settings.

Polypharmacy, a growing concern in healthcare due to multimorbid, aging populations, increases the risk of adverse events and adverse drug reactions in patients. Dose adjustments are often overlooked, limiting knowledge of the likelihood of these adjustments.

Precision medicine is crucial for addressing concomitant diseases and medications, but polypharmacy is particularly observed in hospitalized patients, leading to increased mortality and readmission.

About the study

In the present observational study, researchers performed a comprehensive analysis of 185 million treatment episodes to evaluate the impact of polypharmacy on dose adjustments and identified significant pairs that may influence each other.

The study integrated drug prescription and administrative data from Danish EHRs from 12 public hospitals and was designed to identify significant dose adjustments between specific index drugs and co-medications.

Drug dose modifications in 24 million inpatient prescriptions and admissions covering the period between January 2008 and June 2016 were analyzed in Eastern Denmark (50% of the Danish population), comprising data from all 1,069,873 individuals. The likelihood of dose adjustments when two drugs were administered concomitantly was computed using Bayesian inference analysis and logistic regressions.

Cross-referenced data were obtained from various clinical and bioinformatics drug-drug interaction databases [Danish Medicines Agency database, DrugBank, NLM CV DDI Corpus, ONC-NI, Corpus PK, Cancer Drug Interactions database ONC-HP, Corpus 2011, Twosides, CredibleMeds, Kyoto Encyclopedia of Genes and Genomes (KEGG) Medicus, Corpus 2013, HEP Drug Interactions database, VA-National Drug File-Reference Terminology (VA-NDF-RT) and human immunodeficiency virus (HIV) Drug Interactions database] could be related to pairs associated with dose changes.

The team described associations between polypharmacy and clinical outcomes, diagnoses, and hematological tests and classified them according to their pharmacokinetic properties and drug-drug interaction (DDI) labeling.

To evaluate the clinical implications of the dose-adjusted co-drug pairs, the team analyzed their relationship with diagnoses, blood tests, and outcomes. The co-drug pairs were characterized in relation to the treatment groups of index drugs and the anatomical groups of their co-drugs.

To investigate whether the dose-adjusted co-drug pairs were related to drug-drug interactions, the team cross-referenced dose-adjusted co-drug pairs with established drug-drug interactions, from 15 public-access databases. The team also investigated whether the dose-adjusted co-drug pairs had shared metabolic or transporter activity.

Results

The study identified 77,484 co-drug pairs, with 3,993 likely to be dose-adjusted. Of these pairs, 60% (n=2,412) were related to readmission, mortality, or longer hospital stays while 308 (8%) were associated with reduced kidney function. The team found that over 50% of the identified drug pairs were linked to readmission, mortality, or longer stays, and observed major differences in relation to disease and laboratory tests.

Among 249,379,285 inpatient prescriptions, 902 distinct drugs were prescribed to 50 or more patients, and these drugs were concomitantly administered with up to 857 other drugs. The extent of polypharmacy varied, with a median of drugs ranging from three to eight.

Polypharmacy was prevalent across all ages and correlated positively with age. Dose-adjusted co-drug pairs were dominated by the cardiovascular system, nervous system, and alimentary tract and metabolism system co-medications (above 60%).

Co-drug pairs where index drugs were classified under psycholeptics, psychoanaleptics, antiepileptics, drugs for obstructive respiratory diseases, corticosteroids, dermatological drugs, and antihypertensives had co-drugs in the same anatomic group in more than 40% of dose-adjusted co-drug pairs. Doses for immunosuppressants and antimycotics were often adjusted when co-medicated with anti-infective drugs.

Dose-adjusted co-drug pairs where index groups were classified as antithrombotic agents, agents acting on the renin-angiotensin system, analgesics, antibiotics, anti-inflammatory, and anti-rheumatic agents had the highest odds ratios.

The medical diagnoses that were associated with the majority of dose-adjusted co-drug pairs were cardiometabolic disorders, such as ischemic heart disease, primary hypertension, and dyslipidemia (376 co-drug pairs).

Blood tests that were correlated with the majority of co-drug pairs were cardiometabolic biomarkers, cardiac biomarkers, and coagulation markers. Drug-drug interactions were found in 83% (n=3,297) co-drug pairs. About 50% of co-drug pairs with established drug-drug interactions involved antithrombotic agents, diuretics, beta-blockers, psycholeptics, and psychoanaleptics.

In total, 1,243 concomitantly administered drug pairs (31%) were inhibitors, inducers, or overlapping CYP isozyme substrates, 19% (n=754) pairs with overlapping drug transporters, and 24% (n=948) with genomic variants affecting each other’s transport- and metabolism-related activities.

Overall, the study findings showed that co-drug pairs can help identify drug-drug interactions in polypharmacy based on real-world data. The findings indicated that a simple accumulation of prescribed drugs does not fully capture the complexity of polypharmacy.

Of the dose-adjusted co-drug pairs, 49% were likely to be discontinued during concomitant treatment, highlighting the need for interdisciplinary guidelines. Focusing on drug discontinuation could be an interesting future analysis, as polypharmacy presents additional risks to multimorbid patients.