Visit Website →
← Back to blog

Polypharmacy PGx Reporting: A Lab Build Guide

October 23, 2026
Polypharmacy PGx Reporting: A Lab Build Guide

A pharmacogenomic report for a patient on one drug is a lookup. A report for a patient on eight drugs is a clinical reasoning problem, and the quality of that reasoning depends on three things your lab controls: the accuracy of the medication list behind the report, the way concurrent drugs change a genotype-predicted phenotype, and whether the output reaches the prescriber as something actionable rather than a wall of flags. This guide walks through how to design polypharmacy PGx reporting, including the medication reconciliation step that most report workflows skip.

SignalPGx
signalpgx.com
Make Polypharmacy Reports More Actionable
SignalPGx combines genotype data, medication histories, and clinical evidence in physician-reviewed PGx reports designed for laboratory workflows.
Explore SignalPGx

TL;DR

  • Polypharmacy, commonly defined as five or more concurrent medications, is where pharmacogenomic reporting has the most to offer and the most ways to fail.
  • A PGx recommendation is only as accurate as the medication list it reads. Build reconciliation into the report workflow, not around it.
  • Concurrent inhibitors can shift a genotype-predicted phenotype (phenoconversion), so a genotype-only report can mislead on a complex regimen.
  • Report only actionable drug-gene associations, with consensus terms, and deliver them as discrete data that EHR decision support can consume.
  • Living reanalysis and medical-director review keep multi-drug reports current and defensible.

Table of Contents

Why polypharmacy changes what a PGx report must do

Polypharmacy is usually defined as the simultaneous use of five or more medications. A 2025 paper on pharmacovigilance in primary care cites meta-analysis estimates of how common it is, and ties it to a large burden of adverse drug reactions. Prevalence figures vary by population and definition, so state which one your lab's materials rely on.

The clinical point is subtler than "more drugs, more risk." Medicines optimisation guidance from NICE describes polypharmacy as either appropriate or problematic, depending on whether the prescribing is optimized for complex conditions or whether benefits are not being realized. Pharmacogenomics helps a prescriber tell the two apart, because genetic variability can make a standard drug ineffective or hazardous in a regimen that otherwise looks reasonable. Scotland's 2026 polypharmacy guidance goes further and recommends PGx testing ideally before medications are prescribed, to guide dosing and selection and to reduce the risk of drug interactions and toxicity.

For a laboratory, this means the single-gene, single-drug report template is no longer enough. Your report needs to answer a question the prescriber is really asking: given everything this patient takes, which of these drugs, doses or combinations deserves attention first?

Why polypharmacy changes what a PGx report must do — overview diagram

Where multi-drug exposure concentrates

The overlap between common prescriptions and a small set of genes is large. A Danish register study of drugs with CPIC or DPWG dosing guidelines for CYP2D6 and CYP2C19 mapped how much of national drug use falls on those two enzymes, and argued for introducing PGx testing in polypharmacy patients. A separate analysis of drug-drug-gene interactions found that PPIs and SSRI/SNRIs were among the most critical drug classes for patients with polypharmacy. The practical consequence: a handful of genes will drive most of the actionable content in your multi-drug reports, and your logic for those genes needs to be the most carefully built.

Start with medication reconciliation

Labs often treat the medication list as an input they receive and trust. The evidence says that trust is misplaced. In a pharmacogenomics practice study, Petry and colleagues found that 93% of patients had at least one medication discrepancy in the EHR, and resolving them took an average of 15.8 minutes per patient. A report that confidently cross-references a PGx result against an inaccurate list can be clinically irrelevant, or worse, falsely reassuring.

A companion commentary on the same study, Medication reconciliation to support pharmacogenomics implementation, makes the downstream point: for new prescriptions, decision support embedded in the EHR guides medication and dosing choices, and that only works if results are stored as discrete data and the list the logic reads is correct.

What reconciliation looks like in a lab workflow

You do not have to employ pharmacists to build this. You do have to decide, explicitly, where reconciliation happens and who owns it. A workable design has four parts:

  1. Intake with provenance. Capture each medication with its source (ordering provider, EHR extract, pharmacy fill data, patient-reported) and a timestamp, so the report can say how current the list is.
  2. Normalization. Map free-text and brand names to a standard drug vocabulary so that an entry like a brand-name SSRI and its generic resolve to the same ingredient before logic runs.
  3. Discrepancy flags. Surface duplicates, missing doses, stale entries and conflicts between sources for a human to resolve, rather than silently choosing one.
  4. Stated limitations. Print the medication list date and its sources on the report, and say which drugs were assessed and which were not.

Pharmacist-led approaches to reconciliation have shown that errors can drop when a team iterates on the process, as in this rehab-setting reconciliation project. Your lab's version of that discipline is a defined escalation path: when the list is unreliable, the report says so and the case goes to the ordering clinician before a recommendation is issued.

Phenoconversion: why genotype alone can mislead

In a patient on several drugs, the phenotype predicted from genotype may not be the phenotype the patient actually has. Phenoconversion is the shift that occurs when a co-medication inhibits or induces an enzyme. CYP2D6 is the textbook case: a 2022 expert review notes that it is highly susceptible to inhibition by commonly used medications, so the expected phenotype can differ from the real one.

There is a published method for handling this. A tutorial on integrating CYP2D6 phenoconversion into clinical pharmacogenetics explains that CPIC guidelines refer to an approach in which, for strong inhibitors, the activity score is adjusted to 0, and for weak or moderate inhibitors it is multiplied by 0.5, before converting to a predicted phenotype.

A published case report shows why this matters at the bedside. In a pharmacist-led review, duloxetine inhibited CYP2D6 and converted an intermediate metabolizer into a functional poor metabolizer for co-medications; recommendations that accounted for both the genotype and the interaction were accepted, and the patient's symptoms improved over the following weeks. That is one case, not a trial, so treat it as an illustration of the mechanism rather than proof of an outcome rate.

Reporting phenoconversion responsibly

The decision for your lab is whether to report a genotype-predicted phenotype, an interaction-adjusted phenotype, or both. We recommend showing both, labeled clearly:

  • Genotype-predicted phenotype, using CPIC consensus terms, as the stable, reproducible result.
  • Medication-adjusted consideration, listing the concurrent inhibitors or inducers that were detected on the reconciled list and how they could change the interpretation.
  • A plain statement of confidence, because the adjustment depends on the medication list being complete and on inhibitor strength classifications that your medical director should approve.

Keeping the two layers separate protects you. The genotype result does not change when the patient's regimen does; the adjusted interpretation can, and it should be re-evaluated when it does.

Build the report around actionable content

The technical standard from the American College of Medical Genetics and Genomics for clinical pharmacogenomic testing and reporting sets out what a report should contain: standardized gene nomenclature, variants in HGVS format, all variants and haplotypes detected, stated test limitations, and clear therapeutic recommendations drawn from evidence-based guidelines such as CPIC. It also stresses focusing on actionable associations to avoid alert fatigue.

Polypharmacy raises the stakes on that last point. If a patient takes ten drugs and your report flags every theoretical association, the prescriber stops reading. UK guidance on using pharmacogenomic information in practice says it should be used at prescribing and dispensing where actionable drug-gene associations exist, and a polypharmacy report should follow that principle strictly.

A prioritization pattern for multi-drug reports

TierWhat it containsHow to present it
1. Act nowDrugs on the current list with actionable guideline recommendations for this patient's resultTop of the report, with the specific recommendation and the guideline source
2. Consider with next changeDrugs where the result matters if the dose or agent changes, or interaction-adjusted concernsSecond block, shorter, with the conditions that trigger review
3. Future prescribingDrugs not on the list but relevant to the patient's resultCompact reference table, so it is available at the next prescription
4. No actionable findingDrugs assessed with no guideline actionOne line, so the prescriber knows they were checked

The fourth tier is easy to omit and valuable to include. A prescriber reading a polypharmacy report needs to know which drugs were examined and cleared, not only which were flagged.

Use consensus terms and grade the evidence

Cross-system interoperability depends on shared language. The CPIC consensus paper on standardizing terms for pharmacogenetic test results exists because allele function and phenotype terms varied between laboratories, which makes sharing results across EHRs difficult. Use those terms exactly. On evidence, the EMA's guideline on good pharmacogenomic practice cautions that labs should favor variants with proven functional significance over those only predicted, and that prediction algorithms cannot yet replace functional evidence. For how to map CPIC, PharmGKB and FDA sources into a defensible grade, see our guide to evidence-graded PGx reporting.

Deliver structured data, not only a PDF

A polypharmacy report is most valuable at the moment of the next prescription, which may be months after the lab released it. That requires the result to live in the EHR as discrete data. Research on reporting standards points toward machine-readable, discrete formats for point-of-care decision support, and the CPIC consensus terms are what make those formats interoperable.

The HL7 Genomics Reporting Implementation Guide includes a pharmacogenomic reporting section with a worked walk-through of mapping a PGx report to FHIR. Two practical starting points from our own library:

For developers and health IT teams, the FHIR Genomics IG overview explains how the resources fit together. And if your lab is still deciding how to move from raw genotype calls to phenotype statements, the genotype-to-phenotype reporting guide covers that translation step.

Keep multi-drug reports current

A polypharmacy report has a short shelf life for two reasons. The patient's regimen changes, and the guidelines change. A static PDF handles neither. Build two triggers into your process:

  1. Regimen triggers. When an updated medication list arrives, re-run the interaction and phenoconversion logic against the stored genotype, and issue an addendum if a tier changes.
  2. Evidence triggers. When CPIC, DPWG or FDA labeling content changes for a gene-drug pair in your catalog, identify which released reports are affected and route them for medical-director review.

The second trigger is where manual processes break down. A lab maintaining a spreadsheet of guideline versions will eventually release a report on stale logic. Our CPIC implementation checklist lays out how to version and audit guideline content so that every released report can be traced to the evidence state at the time.

Single-drug examples show how much detail sits behind a good recommendation. See, for instance, our guides on CYP3A5 and tacrolimus dosing, warfarin PGx reporting, HLA-B*15:02 and carbamazepine, and CYP2D6 copy number, which is central to getting the phenoconversion starting point right.

Our perspective: reconciliation is part of the report

We build reporting infrastructure for laboratories, and the pattern we see in multi-drug cases is consistent: labs invest heavily in genotyping accuracy and then hand the interpretation a medication list nobody verified. The report looks polished and the logic is correct, but the input is wrong.

Our view is that a polypharmacy report should be built as a system with four commitments. It treats the medication list as data with provenance and an age. It shows the genotype result and the interaction-adjusted consideration separately. It prioritizes ruthlessly so the prescriber sees what to act on first. And it keeps your medical director in control, with review and an audit trail on every released report, because automation should speed up evidence handling, not replace clinical sign-off.

That is how SignalPGx is designed: automated report generation from genotype and medication data, evidence-graded content, medication intelligence simulation for evaluating drug response and risk, and living reanalysis that updates recommendations as guidelines change, all delivered under your lab's brand and integrated with EHRs through HL7/FHIR. We would rather you ask hard questions about how any platform handles a messy medication list than discover the answer after a report is released.

— Tarek

Putting it into practice with SignalPGx

If you are scoping polypharmacy reporting for your lab, the sequence that tends to work is discovery of your current workflow, a design for how medication data and genotype data meet, and then deployment and integration. We offer those as Discovery & Strategy, System Design, and Deployment & Integration, starting at $7,000, $9,000 and $15,000 respectively as one-off project fees. For labs ready to launch a branded service, White-Label PGx Reporting is available through Pilot, Standard, Scale and Enterprise plans.

A readiness checklist for your lab

Before you commit to a multi-drug report design, check that you can answer yes to these:

  • Can you state where each medication on a report came from, and how old it is?
  • Does your report distinguish genotype-predicted from interaction-adjusted interpretations?
  • Have your medical director and pharmacist reviewers agreed on inhibitor strength classifications?
  • Do you use CPIC consensus terms and HGVS nomenclature consistently?
  • Can the result reach the EHR as discrete data, not only an attachment?
  • Can you identify every released report affected by a guideline change?
  • Does the report list drugs that were assessed with no finding?

Each "no" is a concrete work item, and most can be addressed in the report template and data pipeline before any change to your testing method.

FAQ

What is polypharmacy PGx reporting?

It is pharmacogenomic reporting designed for patients taking multiple concurrent medications, usually five or more. Instead of interpreting one gene-drug pair, the report evaluates the patient's genotype against the whole reconciled medication list, accounts for interactions that may alter enzyme activity, and prioritizes the findings a prescriber should act on.

How does medication reconciliation relate to PGx reports?

The report's recommendations depend on the medication list it reads. In one pharmacogenomics practice study, 93% of patients had at least one discrepancy in their EHR medication list. Reconciling the list before or during report generation is what keeps the recommendations clinically relevant.

Does a genotype-only report work for patients on many drugs?

It can miss important context. Co-medications that inhibit an enzyme such as CYP2D6 can change the effective phenotype, a process called phenoconversion. A report that presents the genotype-predicted result alongside a clearly labeled medication-adjusted consideration gives the prescriber a more complete picture.

How do we avoid alert fatigue in a multi-drug report?

Limit the report's headline content to actionable drug-gene associations backed by guidelines such as CPIC, and rank them by urgency. Move lower-priority information into a reference table, and list drugs with no actionable finding on one line so the prescriber knows they were checked.

Sources