From periodic medication review to proactive medication surveillance

Patients
pharmacodynamics concept

Medication risk does not stay fixed after a prescription is written. As regimens, physiology, and clinical information change, medication review needs to become more responsive to those changes.

A medication can be completely appropriate when it is prescribed and become a problem later. Another drug may be added that changes its metabolism. Kidney or liver function may decline. New symptoms may emerge that are medication related. A pharmacogenomic result, informing about the patient’s inherited ability to respond to a medication, may already be sitting in the health record, largely irrelevant at the time it was generated, but suddenly important when a new therapy is started. That is the part traditional medication review does not always capture well.

Medication review is still often treated as a point-in-time activity, focused on medication reconciliation, adherence, and whether each drug remains appropriate. Those questions remain important, but they are no longer enough. As treatment regimens become more complex, the bigger issue is whether something has changed that alters the safety or effectiveness of the regimen.

That shift suggests a different model, one built less around periodic review alone and more around ongoing medication-risk surveillance, with technology helping identify meaningful changes and clinicians focusing their attention where it is most needed.

Medication appropriateness changes over time

Medication reconciliation answers a basic question: what is the patient taking? Medication review goes further: are those medications still needed, safe and effective? But there is another question that deserves equal attention: what has changed since the last review?

Medication risk can shift for many reasons. A drug is added or stopped. A dose changes. Renal or hepatic function worsens. A patient develops a new condition. Care moves from one setting to another. New laboratory or genomic information becomes available. New symptoms can also complicate the picture, particularly when they are attributed to disease progression or a new condition, rather than recognised as a possible medication-related effect.

Drug-drug interactions, or DDIs, show how often the medication environment itself can create a new risk. A 2025 analysis of the FDA Adverse Event Reporting System (FAERS) identified 167,065 reports involving DDIs submitted before April 2024. Warfarin, aspirin, sertraline, tacrolimus, and simvastatin were among the medications most frequently reported. FAERS is a spontaneous reporting system, so this data cannot establish incidents or causality. Still, they are a reminder that interaction risk remains a persistent part of medication management.

When drug-drug interactions and genetics overlap

The picture gets more complicated when inherited genetic variation is part of the story. Simvastatin, a widely prescribed cholesterol lowering medication, is a good example. Variants affecting SLCO1B1, a transporter involved in its liver uptake, can increase simvastatin blood concentrations and raise susceptibility to myopathy. Current clinical pharmacogenetics implementation consortium guidance recommends an alternative medication for patients with poor SLCO1B1 function and recommends considering an alternative or limiting the simvastatin dose, for decreased function phenotypes. The same guidance also notes that DDIs can further increase myopathy risk.

Now, imagine a patient who starts simvastatin without difficulty. Nothing about the original prescription was necessarily wrong. Then another medication is added, that changes simvastatin pharmacokinetics. The risk picture is different if the patient also has decreased SLCO1B1 function; there is another layer of susceptibility.

A DDI review catches one part of that story. A pharmacogenomic review catches another. What matters clinically is how those factors come together in the patient at that point in time.

The broader point is simple. Medication review should not stop at the question of whether a drug is still indicated. It should also ask whether the patient's current biology and medication environment have changed the expected response.

A genotype is stable, but the functional phenotype may not be

One of the clearest examples comes from a study of more than 2.4 million Veterans Health Administration pharmacy users who received an opioid prescription. Researchers looked at CYP2D6-metabolised opioids, including codeine, hydrocodone, and tramadol, together with interacting medications. CYP2D6 is a key metabolising enzyme for these opioids.

Twenty-eight percent of patients were also prescribed antidepressants that interact with the CYP2D6 enzyme. Overall, the investigators estimated that 21.6% were at elevated risk of an undesirable opioid response based on predicted phenotype from their pharmacogenomic test results and drug-drug interactions. Particularly striking was the finding that 12.8% were normal or intermediate metabolisers who were also taking a CYP2D6 inhibitor, potentially shifting their functional phenotype toward poor metabolism, putting them at risk of uncontrolled pain.

That distinction is easy to miss. A person’s inherited genotype does not change. But the way that genotype is expressed clinically can change as medications are started or stopped. A pharmacogenomic result, then, should not necessarily be interpreted once and filed away. Its relevance can change over the course of a patient's life.

Medication review needs to become more dynamic

The practical answer is not to ask pharmacists or physicians to re-review every medication relationship every day. A more realistic approach is ongoing surveillance with targeted human review. Technology can watch for changes that matter, such as a new medication, a dose adjustment, a new pharmacogenomic result, a change in renal or hepatic function, a transition of care, or a meaningful shift in the overall regimen. When one of those changes occurs, the system can reassess relevant drug-drug, drug-gene, and clinical relationships, and determine whether the case deserves attention.

From there, a pharmacist or other qualified clinician reviews the finding, decides whether action is needed, and the patient’s medication profile is reassessed as new information becomes available.

The answer is not more alerts

Clinicians already work in an environment saturated with alerts. Adding more warnings is unlikely to solve the problem. The better question is whether we can identify the relatively small number of medication changes that are most likely to affect a treatment decision. Clinical decision support, rule-based analytics, and AI-assisted tools can help identify and prioritise changes that warrant clinical review. Their value is not in replacing clinical judgement. It is in reducing the search burden and helping clinicians focus on the cases that deserve a closer look.

Some patient groups are obvious places to start. These include patients with polypharmacy, multiple prescribers, high-risk medications, impaired renal or hepatic function, multiple chronic conditions, frequent transitions between care settings, or existing pharmacogenomic results that may become relevant when treatment changes.

Healthcare organisations do not need to build this capability for every patient at once. A risk-based approach can begin with populations in whom medication-related problems are both more likely and more consequential.

Building that capability takes more than an interaction database. It requires integrated medication and clinical data, longitudinal access to genomic results, sensible prioritisation rules, and a clear pathway for a pharmacist or other qualified clinician to review the finding and act on it when needed.

Medication lists are snapshots; medication risk is not

The next step in medication review is not simply to perform reconciliation more often. It is to recognise that medication safety and effectiveness can change as the patient changes. The goal is not to review every medication every day. It is to recognise when something has changed enough to make another review worthwhile.

About the author

Houda Hachad is a precision medicine executive with over 20 years of experience bridging clinical pharmacology, genomic, informatics, and healthcare innovation. As Aranscia’s vice president of clinical operations, she leads initiatives that expand turnkey precision-medicine solutions. Her work focuses on aligning and integrating pharmacogenomic testing, clinical decision support, and workflow optimisation across health organisations, laboratories, and payer networks. She is actively engaged in shaping the pharmacogenomics ecosystem through leadership and participation in CPIC, PharmVar, AMP PGx, PGRN, and the STRIPE collaborative community. Recipient of the 2024 STRIPE Double Helix Award and 2022 PharmaVoice voice 100 “Tech Wizards” honour, she champions practical, equitable genomic adoption in healthcare.

Image
Houda Hachad
profile mask

Houda Hachad