Home › Test Catalogue
Test Catalogue
Three disciplines. One building.
Toxicology, molecular infectious disease, and genomic sequencing — every assay validated, performed, and reported by Prime Path staff in Denver. Choose a discipline below, or read how these panels were designed and validated before they accepted a single patient specimen.
Urine Drug Testing
Toxicology
Presumptive immunoassay screening and definitive LC-MS/MS confirmation, with pH, specific gravity, and creatinine validity testing on every specimen. Built for pain management, behavioral health, and compliance monitoring.
View panelsMolecular PCR
Infectious Disease
Seven real-time multiplex PCR panels covering urinary, respiratory, wound, nail, ocular, ENT, and sexually transmitted infection. Organism identity and antimicrobial resistance markers on the same report.
View panelsNext-Generation Sequencing
Genetic & Genomic
Twelve targeted NGS panels from a single non-invasive buccal swab — hereditary cancer, pharmacogenomics, cardiogenetics, neurology, and more. Sequencing and bioinformatics never leave the building.
View panelsThe evidence base
What earns a gene a place on the panel.
A panel is a claim. Every gene on it asserts that variation in that gene changes what a clinician should do next. Most of the genes that could be added to a panel do not clear that bar.
Gene–disease validity is not a matter of opinion. It is curated, versioned, and publicly auditable through resources such as ClinGen, the Gene Curation Coalition (GenCC), and Genomics England PanelApp, each of which grades the strength of an asserted relationship on a defined scale. Prime Path designs its genomic panels against those classifications rather than against the maximum number of genes a sequencing chemistry can technically capture.
Definitive / Strong
The gene–disease relationship is upheld by multiple independent lines of evidence and has not been contradicted. These genes belong on a diagnostic panel.
Moderate
Evidence is real but thinner. Included only where the clinical question makes a moderate-validity gene actionable, and flagged as such in interpretation.
Limited / Disputed
Reported associations that have not held up, or that active work contradicts. A limited-validity gene inflates a panel’s gene count and its rate of uninterpretable findings. We leave them out.
Refuted
The association has been overturned. Ordering a gene in this tier does not produce a negative result. It produces noise.
The commercial incentive runs the other way. A 500-gene panel photographs better than a 60-gene panel, prices higher, and is materially worse for the patient: every low-validity gene added to an assay raises the probability of returning a variant of uncertain significance, and a VUS is a finding that cannot be acted on but cannot be unseen either. It generates surveillance, anxiety, cascade testing of relatives, and occasionally surgery, on the strength of evidence that does not support any of it.
When new evidence reclassifies a gene — upward or downward — the panel changes. That is what a curated panel means, and it is why we publish the gene content of each assay on its brochure rather than describing it as “comprehensive.”
Panel design
Why targeted, and why customised.
Prime Path runs targeted next-generation sequencing. We do not offer whole exome or whole genome sequencing, and the omission is deliberate.
| Approach | What it reads | Depth of coverage | Clinical cost |
|---|---|---|---|
| Targeted panel | A curated gene set chosen for one clinical question | High and uniform across every targeted base | Low VUS burden; clear indication for use |
| Whole exome | ~1–2% of the genome; protein-coding regions | Lower and uneven; coverage gaps in GC-rich exons | Substantial VUS and secondary-finding burden |
| Whole genome | Effectively everything, coding and non-coding | Lower still per base at comparable cost | Interpretation burden exceeds most clinical questions |
Depth is the trade. A targeted panel concentrates sequencing reads onto the bases that answer the question, which is what allows confident variant calling at the coverage thresholds a diagnostic assay requires. Spread the same reads across an exome and per-base depth falls; spread them across a genome and it falls further. For a hereditary cancer question or an arrhythmia question, the exome is mostly reading DNA nobody asked about, less well.
Customisation happens at the level of the clinical question, not the catalogue. Our cardiogenetics panel separates the structural drivers of cardiomyopathy from the channelopathy genes behind long QT and Brugada, because those two findings lead to different surveillance and different device decisions. Our neurological panel is not one list — it is epilepsy, neuropathy, ataxia, and early-onset dementia, carved so a clinician can order the subpanel their patient actually presents with.
- Panels are scoped to a presentation, not to a marketing category.
- Subpanel carve-outs exist so a single-gene question is not billed as a 164-gene panel.
- Gene content is published on each brochure, not summarised as “comprehensive.”
- Secondary findings are not reported unless the ordering provider has requested them.
Analytical validation
“Validated” is a legal term, not an adjective.
Under 42 CFR 493.1253, a laboratory that introduces a test not cleared by the FDA, or modifies one that was, must establish its performance characteristics before reporting a patient result. Not the manufacturer’s. Its own.
Design
Gene content, targets, and clinical claim are fixed before a reagent is ordered.
Optimisation
Chemistry, thermal profile, and bioinformatics thresholds are tuned on known material.
Establishment
Performance characteristics are measured and documented against 42 CFR 493.1253.
Verification
Independent confirmation on blinded and previously characterised specimens.
Release
Director signs. Only then does the assay accept a patient specimen.
That obligation attaches to the instrument in this building, the reagent lot on this bench, and the staff who will run it. A validation performed by a vendor in another state, on another analyser, is a starting point and nothing more. What follows is what we establish and record for every assay before it goes live.
| Performance characteristic | The question it answers |
|---|---|
| Accuracy | Does the assay return the right answer against a reference method or characterised material? |
| Precision | Does it return the same answer within a run, between runs, and between days? |
| Analytical sensitivity | What is the lowest concentration — limit of detection — the assay reliably calls? |
| Analytical specificity | What interferes? Cross-reactivity, interfering substances, and near-neighbour organisms or alleles. |
| Reportable range | Between what limits is a quantitative result trustworthy? |
| Reference intervals | What does a result mean in the population this laboratory serves? |
Beyond release, an assay is not left alone. Every analytical run carries internal quality controls, and a control outside its established limits stops the run rather than appending a caveat to a report. Reagent lots are verified before use. Instruments are maintained and calibration verified on documented schedules. Personnel competency is assessed and recorded, because a validated method executed by an unassessed operator is not a validated result. Blinded proficiency-testing specimens arrive from an external programme and are graded against peer laboratories running the same method — the closest thing clinical diagnostics has to an unannounced examination.
Depth and confidence
What sequencing depth actually buys.
Every base in a sequencing run is read some number of times. The number is not a vanity metric — it is the difference between a variant you can report and a variant you cannot.
Schematic. Illustrates why targeted panels concentrate reads: confidence in a variant call rises steeply with depth, then plateaus. Curve shape is illustrative and is not Prime Path performance data.
At low depth, a heterozygous variant can be missed entirely because the reads that carried it never appeared, and a sequencing artefact can masquerade as a real call because nothing contradicted it. As depth rises, both error modes collapse: the real variant is seen repeatedly across independent reads, and the artefact is outvoted. Above the plateau, additional reads buy very little.
This is the whole argument for targeting. A finite number of reads comes off the flow cell. Spend them on 60 curated genes and every base clears the threshold comfortably. Spend the same reads across 20,000 genes and a meaningful fraction of clinically important exons — often the GC-rich first exons where pathogenic variants cluster — drop below it. Those bases still appear on the report as “no variant detected.” They were never adequately read.
Our panels specify a minimum depth at every targeted base, and any base falling below it is reported as a coverage gap rather than a negative result. A physician is entitled to know the difference between “we looked and found nothing” and “we could not see.”
Coverage & claims
Panels built against coverage policy, not against catalogues.
A test that is clinically sound and reimbursement-blind still lands on your practice as a denial, an appeal, and a patient bill nobody expected.
We build our urine drug screening and infectious disease panels against current CMS coverage policy — against what a payer will recognise as medically necessary for a stated indication, rather than against what a panel could theoretically include. The practical effect is narrower panels, cleaner claims, and less of your staff’s time spent on appeals.
- Indication drives content. The requisition captures the clinical indication and ICD-10 code, because the panel that is defensible for a symptomatic patient is not the panel that is defensible for a screening encounter.
- Definitive testing is ordered, not reflexed silently. A presumptive positive is confirmed by LC-MS/MS when confirmation will inform a decision — and we tell you when a reflex adds a charge.
- Subpanels exist so single-gene questions bill as single-gene questions. Ordering a 164-gene panel to answer a one-gene question is a coverage problem before it is a clinical one.
- Prices are published. Self-pay pricing sits on this website, in a PDF, with the code and the number.
None of this is generosity. A laboratory that designs for reimbursement designs for restraint, and restraint is the same discipline that keeps a panel free of limited-validity genes and a report free of findings nobody can act on. The incentives happen to align. Where they do not, the clinical question wins, and we will tell you before you ship the specimen rather than after the denial arrives.