Therapeutic Drug Monitoring and Overdose Testing

Absorption determines how much of an oral dose reaches the circulation. Dissolution, solubility in gastrointestinal fluid, and passive diffusion of the nonionized form all limit uptake, and the small intestine absorbs most oral drugs because of its surface area and blood flow. Intravenous administration bypasses absorption and gives complete, predictable bioavailability, which is why an intravenous drug reaches its highest concentration at the end of the infusion.

Distribution moves drug between blood and tissue. The volume of distribution, Vd, is the amount of drug in the body divided by the plasma concentration. It is large for lipophilic drugs that partition into fat and nerve tissue and small for drugs confined largely to plasma, including ionized and heavily protein-bound compounds. Vd is an apparent volume that corresponds to no anatomic space, and after an intravenous bolus it can be estimated from the dose and the concentration extrapolated back to time zero. Only unbound drug crosses membranes, engages receptors, and is available for metabolism and glomerular filtration. Protein binding shifts with inflammation, malignancy, pregnancy, uremia, hepatic disease, hypoalbuminemia, and competing drugs, so the active fraction can change when the total concentration does not.^{11 Rifai N, et al. Tietz Textbook of Laboratory Medicine. 2023.}

Metabolism converts hydrophobic compounds into excretable ones, mainly through the hepatic cytochrome P450 enzymes, historically described as the mixed-function oxidase system. Phase I reactions oxidize, reduce, or hydrolyze the parent drug to expose or add a functional group. Phase II reactions conjugate the parent or its phase I product by glucuronidation, sulfation, acetylation, glutathione conjugation, or amino-acid conjugation. Enzyme induction, as by carbamazepine, phenytoin, phenobarbital, and rifampin, accelerates clearance and shortens half-life. Inhibition by azole antifungals, macrolides, and other agents slows clearance and lengthens half-life. Cirrhosis and other hepatic disease reduce metabolic capacity. First-pass metabolism removes part of an orally absorbed dose during its initial transit through gut wall and liver, so oral and intravenous doses of the same drug are not interchangeable.

Most drugs are eliminated by first-order kinetics, a constant fractional decline per unit time:

where k is the elimination rate constant and t½ the half-life. A few drugs, especially phenytoin, saturate their metabolizing enzymes within the therapeutic range and shift toward zero-order elimination, so a small dose increase can produce a disproportionate concentration rise. On a fixed regimen the concentration approaches steady state after roughly five half-lives, when input and elimination balance. Five half-lives is an approximation, and concentrations drawn earlier still guide loading, suspected toxicity, deteriorating renal or hepatic function, and exposure estimation.^{11 Rifai N, et al. Tietz Textbook of Laboratory Medicine. 2023.}

The therapeutic index, conventionally TD₅₀/ED₅₀, expresses how much separation a drug allows between an effective and a toxic exposure. The toxicology lesson covers those dose-response terms.

Peak, trough, and exposure. A trough is the lowest concentration in a dosing interval, collected immediately before the next dose. A peak is the highest, but its clock time depends on route, formulation, infusion duration, and distribution, so the protocol defines each peak time. No fixed number of minutes applies. Area under the concentration-time curve, AUC, describes total exposure across an interval and requires either two timed concentrations with first-order equations or Bayesian software using one or two concentrations against a population model. Peak, trough, and AUC each describe a different aspect of exposure, and none can be determined from a specimen whose collection time is unknown.

k = ln(30 ÷ 20) ÷ 6 h = 0.0676/h

t½ = 0.693 ÷ 0.0676/h = 10.3 h

Predicted 21:00 concentration = 20 µg/mL × e^{(−0.0676 × 6)} = 13.3 µg/mL

The prediction assumes first-order elimination continues. The calculation characterizes one patient's elimination and supports an exposure estimate. It does not by itself set a dose. For serious methicillin-resistant Staphylococcus aureus (MRSA) infection, the commonly used AUC over 24 hours (AUC₂₄) target is 400–600 mg·h/L when the broth-microdilution minimum inhibitory concentration (MIC) is assumed to be 1 mg/L. AUC₂₄/MIC is a separate ratio: dividing mg·h/L by mg/L leaves hours. The responsible clinical service selects the regimen.^{22 Rybak MJ, et al. Therapeutic monitoring of vancomycin for serious… 2020.}

Collection time is the dominant preanalytic variable in therapeutic drug monitoring, and it is the one the laboratory most often influences. Every request needs the drug, the dose and route, the time the last dose was given or the infusion ended, and the time of the draw. A specimen with no recorded collection time cannot be compared with a peak, trough, or exposure target. Report the concentration with the missing information stated, or recollect, so that no target range is applied to an unknown time point.^{33 Clinical and Laboratory Standards Institute. Toxicology and Drug Testing in the Medical Laboratory. 2021.}

Draw a trough immediately before the next scheduled dose when the protocol calls for a trough. Post-dose sampling times are protocol-specific and differ between drugs, and between dosing strategies for the same drug, so follow the local monitoring protocol for aminoglycoside and vancomycin sampling.

Tube and specimen requirements vary by assay, so collect what your laboratory specifies. This includes calcium-chelating anticoagulants, which some drug assays accept and others do not.

• Gel-barrier separator tubes adsorb lipophilic drugs onto the gel, and the loss grows with contact time. Phenytoin is the classic example, and the measured concentration falls. Collect these drugs in a plain tube without gel, or separate promptly.^{44 Mayo Clinic Laboratories. Test catalog: phenytoin, total and free, digoxin, serum… 2025.}
• Lithium is collected in serum. A lithium-heparin tube contaminates the specimen and grossly raises the result.
• Tacrolimus, cyclosporine, sirolimus, and everolimus partition heavily into erythrocytes and are measured in EDTA whole blood, which is required. Serum or plasma markedly underestimates the circulating concentration.^{55 ARUP Laboratories. Test directory.}
• Mycophenolic acid stays extracellular and is measured in plasma or serum, with the individual assay defining which. The assay's specimen requirement applies even when another drug in the same class uses a different specimen.^{55 ARUP Laboratories. Test directory.}
• Aminoglycosides and beta-lactams inactivate each other in vitro, and the reaction continues in a collected specimen, so a delayed aminoglycoside assay reads falsely low in a patient receiving both.^{66 US Food and Drug Administration. Amikacin sulfate injection prescribing information.} In one serum mixing study the effect was marked with penicillin G and ampicillin, moderate with nafcillin and cefoxitin, and slight with cefazolin. Tobramycin was the most reactive aminoglycoside and amikacin the most stable.^{2323 Tindula RJ, et al. Aminoglycoside inactivation by penicillins and cephalosporins… 1983.} Assay promptly, freeze the specimen, or use a validated beta-lactamase treatment.^{66 US Food and Drug Administration. Amikacin sulfate injection prescribing information.}

A specimen drawn from a line or lumen used to infuse the same drug is the most serious collection error in drug monitoring. Residual drug in the catheter produces results many times the true concentration, so rule out a line draw before anyone interprets an implausibly high immunosuppressant or vancomycin result clinically.

Routine assays measure total drug, bound plus unbound. For heavily protein-bound drugs the total can misrepresent the active fraction whenever binding changes. Phenytoin is about 90% bound to albumin, so the free fraction near 10% accounts for all the pharmacologic activity. Hypoalbuminemia, uremia with retained organic acids competing for binding sites, pregnancy, and displacing drugs such as valproate all raise the free fraction. A patient can then show toxicity when the total concentration is within the reference range. Free measurement, by ultrafiltration or equilibrium dialysis at 37 °C, is the appropriate test in those settings.^{44 Mayo Clinic Laboratories. Test catalog: phenytoin, total and free, digoxin, serum… 2025.}

Albumin-based correction equations estimate what the total concentration would be at normal albumin. They assume otherwise normal binding, perform poorly in uremia and critical illness, and cannot substitute for a measured free concentration when the question is whether the active fraction is toxic. Valproic acid behaves similarly and less predictably, because its binding saturates within the therapeutic range: the free fraction rises disproportionately as the total rises, and again in hypoalbuminemia and renal failure.^{11 Rifai N, et al. Tietz Textbook of Laboratory Medicine. 2023.}

Every target below is method-, indication-, and protocol-dependent. A published range is a guide to interpretation and cannot be used alone as a decision threshold. Some patients respond, or show toxicity, at concentrations outside it, and each laboratory reports against the range it has verified for its own assay.

Antimicrobials

Vancomycin monitoring changed strategy in 2020. Trough-only targeting of 15 to 20 mg/L, the previous standard, is no longer recommended for serious MRSA infection, because a trough is an unreliable surrogate for total exposure. Exposure is now estimated either from two timed concentrations, a post-distribution level one to two hours after the infusion ends plus a trough in the same interval, or by Bayesian software using one or two concentrations with at least one trough. Evidence was judged insufficient to require either approach in noninvasive MRSA or non-MRSA infection, so those patients are monitored under local protocol. For the laboratory, this means infusion start and stop times, dose, and draw times must be accurate, and turnaround must fit a target-attainment window inside the first 24 to 48 hours of therapy.^{22 Rybak MJ, et al. Therapeutic monitoring of vancomycin for serious… 2020.}

Aminoglycoside sampling depends entirely on the dosing strategy, and applying one strategy's targets to the other gives a wrong interpretation.

Some drugs in older monitoring lists are less commonly measured or are monitored under specialty protocols. Availability, referral arrangements, and the value of a concentration depend on the drug, formulation, clinical question, and laboratory.

Older method lists include fluorescence-polarization immunoassay, radioimmunoassay, and gas-liquid chromatography. Identify the laboratory's current method and its validation before interpreting its results.

Most therapeutic drugs are quantified by competitive immunoassay, in which drug in the specimen competes with a labeled drug or a drug-coated particle for a limited quantity of antibody. Signal direction depends on the format. Labeled-antigen and agglutination-inhibition formats give a signal that falls as concentration rises. Homogeneous enzyme formats give the opposite response. In enzyme-multiplied immunoassay and cloned enzyme donor immunoassay, specimen drug frees the enzyme label to become active, so absorbance rises with concentration. Formats in current US routine use include enzyme-multiplied immunoassay, cloned enzyme donor immunoassay, chemiluminescent and electrochemiluminescent microparticle immunoassays, kinetic interaction of microparticles in solution, particle-enhanced turbidimetric inhibition, and antibody-conjugated magnetic immunoassay. The immunology lesson explains how these assay formats work.^{11 Rifai N, et al. Tietz Textbook of Laboratory Medicine. 2023.}

The main limitation of immunoassay is specificity. An antibody raised against a parent drug cannot reliably exclude an active metabolite, an inactive metabolite, a prodrug, or a structurally related compound, and the magnitude of that cross-reactivity is specific to the individual reagent. Published values for immunosuppressant metabolites span a few percent to essentially complete: one current tacrolimus method sheet reports 70% cross-reactivity for the M-II metabolite and none detectable for three others. Another platform reports near-complete cross-reactivity for a different metabolite.^{1111 Roche Diagnostics. Elecsys immunosuppressant drug method sheets.} Common examples are carbamazepine-epoxide in carbamazepine assays, fosphenytoin in phenytoin assays, immunosuppressant metabolites in whole-blood assays, sirolimus and everolimus in each other's assays, DAMPA in methotrexate assays after glucarpidase, and digoxin-like immunoreactive substances and antibody Fab fragments in digoxin assays.

LC-MS/MS resolves parent drug from metabolites by mass and is the routine method wherever cross-reactivity is clinically unacceptable, especially the immunosuppressants, along with the newer anticonvulsants, the antifungals, and analytes with no commercial immunoassay.^{55 ARUP Laboratories. Test directory.} It has its own failure modes. Matrix effects can suppress or enhance ionization, isobaric and isomeric compounds can interfere, and carryover can contaminate a later injection. Suitable isotope-labeled internal standards help compensate for recovery and matrix effects. Establish selectivity through chromatographic and mass-spectrometric validation. Assess carryover with blanks after high-concentration specimens and validated wash or sequence controls.^{2222 International Council for Harmonisation. M10: Bioanalytical Method Validation and Study Sample Analysis. 2022.} Batch scheduling and specialist staffing lengthen turnaround, which matters when the result is meant to inform dosing in the first two days of therapy. Most clinical LC-MS/MS drug assays are laboratory-developed tests, validated under the Clinical Laboratory Improvement Amendments (CLIA) and inspected against College of American Pathologists (CAP) checklist requirements, without FDA clearance.

Standardization and platform differences. Therapeutic drug monitoring is unevenly standardized. Phenytoin, valproate, and the other classic anticonvulsants agree reasonably well between methods. The immunosuppressants do not: method comparisons in current package inserts show slopes above 1.0 against LC-MS/MS, with the size of the positive bias varying across reagents, so the general observation that immunoassay reads high describes a tendency and is never used as a correction factor.^{1111 Roche Diagnostics. Elecsys immunosuppressant drug method sheets.} Candidate reference measurement procedures now exist for some of these analytes, mycophenolic acid among them. A candidate reference procedure does not yet make routine results interchangeable. In practice, keep a patient's serial results on one method, validate a program's target range against the method in use, and communicate any method or platform change to the clinical services that use the results.

These assays sit under the same CLIA quality-system requirements as the rest of chemistry, and the laboratory operations area covers those general rules.

Reportable range matters more than usual, because overdose concentrations routinely exceed the analytical measurement range. Dilution protocols must be validated for the specific assay, and current immunosuppressant method sheets restrict manual dilution to narrow conditions, so an out-of-range immunoassay result may need a different method. Functional sensitivity limits interpretation at the other end: several current whole-blood immunoassays quantify no lower than 1 to 1.5 ng/mL, which is inside the range where a transplant trough can matter. Calibration verification and lot-to-lot comparison are important because a small calibration shift changes the apparent trough for every transplant patient tested. Proficiency testing is the only routine means of assessing between-method bias, and a persistent peer-group offset warrants investigation.

Result verification for these analytes is mostly a check that the specimen matches the request. A concentration far above anything the dose could produce suggests a line draw or a contaminated tube, so rule these out before attributing the result to the patient. A concentration near zero in an adherent patient suggests a missed dose, a draw at the wrong point in the interval, or the wrong tube. A delta check against the previous result is informative when the regimen has not changed and much less so during titration.

The laboratory's medical director sets critical and alert limits for drug concentrations in the institution's policy. Lists commonly include lithium, digoxin, phenytoin, valproate, carbamazepine, phenobarbital, theophylline where still offered, acetaminophen, and salicylate. CLIA requires the laboratory to immediately alert the individual or entity requesting the test, and where applicable the person responsible for using the result, whenever a result indicates an imminently life-threatening condition or meets the laboratory's alert criteria. Document the date, time, result, and person notified.^{1212 Centers for Medicare & Medicaid Services. 42 CFR § 493.1291(g).} Read-back of the value and patient identifiers, defined escalation when the first contact fails, and review of delayed notifications are part of the same policy.^{1313 Clinical and Laboratory Standards Institute. Management of Critical- and Significant-Risk Results. 2015.}

A concentration reported without its collection time can be misinterpreted, so flag a mistimed specimen on the report. When a result is implausible or an interference is suspected, state the suspected interference and the alternative method. The laboratory advises on specimen, timing, and assay questions, and dose decisions are made by the prescriber.

Acetaminophen and salicylate are the main reason a general hospital laboratory offers drug assays around the clock. Interpretation of both depends on timing. Their reporting units are easily confused, and the patient often cannot give a reliable history.

Acetaminophen. At therapeutic doses acetaminophen is conjugated to harmless glucuronide and sulfate, with a small fraction oxidized by cytochrome P450 to a reactive intermediate that glutathione neutralizes. Overdose saturates conjugation, exhausts hepatic glutathione, and lets the intermediate accumulate and kill hepatocytes. The early clinical picture and the early laboratory panel can appear normal. Aminotransferases typically begin rising 24 to 72 hours after ingestion and peak around 72 to 96 hours, so interpretation depends on a timed concentration.

For an acute ingestion at a known time, a concentration drawn no earlier than 4 hours after ingestion is plotted on the Rumack-Matthew nomogram. The 2023 US and Canada consensus counts an ingestion as acute when all of it was taken within a period shorter than 24 hours, as one dose or several. The US treatment line passes through 150 µg/mL at 4 hours and falls with a 4-hour half-life, so it sits at 75 µg/mL at 8 hours, 37.5 µg/mL at 12 hours, and near 4.7 µg/mL at 24 hours. That line lies 25% below the original 200 µg/mL line as a safety margin, and it is valid only from 4 to 24 hours after an acute ingestion at a known time. A concentration drawn before 4 hours cannot be plotted on the nomogram, because absorption and distribution are incomplete, and an early result is a common source of false reassurance. Unknown-time ingestions and repeated supratherapeutic ingestion, spread over more than 24 hours, require separate assessment. High-risk acute ingestion at a known time still uses the applicable timed-concentration pathway, including the high-risk nomogram line, with additional specialist management. For repeated supratherapeutic ingestion, the 2023 US and Canada consensus treats a concentration above 20 µg/mL or abnormal aminotransferases as a clinical action threshold. The laboratory reports and communicates the timed concentration and liver tests. The poison service or treating team determines antidote use.^{1414 Dart RC, et al. Management of acetaminophen poisoning in the US and Canada. 2023.}

Reporting units are a common source of error. US laboratories report µg/mL, equivalently mg/L. SI reporting uses µmol/L. Multiplying µg/mL by 6.62 gives µmol/L, so the 4-hour treatment line of 150 µg/mL is close to 1,000 µmol/L. Routine measurement uses enzymatic colorimetric assays, immunoassay, or chromatography. Enzymatic assays carry documented positive interference at high bilirubin concentrations, with limits stated in the method insert, and chromatography is the reference method and the alternative when interference is suspected.

Salicylate. Salicylates inhibit cyclooxygenase, reducing thromboxane and prostaglandin synthesis, which produces platelet dysfunction and gastric mucosal injury, and their use in children with viral illness is epidemiologically linked to Reye syndrome. Overdose produces a mixed acid-base disturbance characteristic enough to suggest the diagnosis before the concentration is reported. Salicylate directly stimulates the respiratory center, producing respiratory alkalosis. It also uncouples oxidative phosphorylation and inhibits Krebs-cycle enzymes, which generates lactate, and drives lipolysis, which generates ketoacids. Adults usually present with the mixed picture, and young children move to a predominantly high-anion-gap metabolic acidosis earlier. The acid-base lesson covers the interpretation.^{11 Rifai N, et al. Tietz Textbook of Laboratory Medicine. 2023.}

Therapeutic concentrations are about 3 to 10 mg/dL for analgesia and 15 to 30 mg/dL for anti-inflammatory use. Toxicity generally begins above roughly 30 to 40 mg/dL after acute ingestion and is severe above 100 mg/dL, but a single value cannot be interpreted without acuity, pH, and the clinical trajectory. Chronic salicylate toxicity is dangerous at concentrations that would be unremarkable acutely, because tissue distribution, more than the plasma concentration, determines central nervous system exposure. Serial concentrations are collected until clearly falling and continued absorption is unlikely. A falling plasma concentration can be misleading. As pH falls, more salicylate exists in the nonionized form and enters the central nervous system while the plasma value declines.^{1515 Dargan PI, et al. An evidence based flowchart to guide the management of acute… 2002.}

The Done nomogram, which purported to grade salicylate toxicity from one concentration and a time since ingestion, is unreliable in both acute and chronic poisoning and should not be used. Unlike acetaminophen, salicylate has no valid single-concentration nomogram.^{1515 Dargan PI, et al. An evidence based flowchart to guide the management of acute… 2002.}

The Trinder reaction, in which ferric ion forms a colored complex with salicylate, remains a common quantitative method. Enzymatic and chromatographic methods give better specificity. Trinder-type chemistry reacts with structurally related compounds, and diflunisal and salicylamide are documented positive interferents. Qualitative urine ferric chloride and Trinder spot tests have good sensitivity and limited specificity. A negative spot test still requires a quantitative serum concentration. Reporting units are a common source of error: the same result is 30 mg/dL, 300 mg/L, or about 2.2 mmol/L, and a tenfold unit error is the classic critical-value error for this analyte.

Urine immunoassay screening followed by chromatographic or mass-spectrometric confirmation is the standard two-tier design. A positive screen means that antibody reactivity exceeded the cutoff. The antibody recognizes a drug class and may cross-react with other compounds, so the result remains presumptive until confirmatory testing, such as mass spectrometry, identifies the specific drug. A negative screen means the result is below the cutoff. The analyte may still be present. Screen and confirmation cutoffs differ for the same class, so a cannabinoid immunoassay screening at 50 ng/mL is confirmed at 15 ng/mL for the carboxy metabolite. Report presumptive positives with the method, the cutoff, and the limitation stated. Current US guidance for emergency department testing recommends dropping the tricyclic antidepressant and propoxyphene screens from that menu, because they add little and propoxyphene left the US market more than a decade ago.^{1616 Snozek CLH, et al. ADLM guidance document on laboratory testing for drugs of misuse… 2026.}

Clinical and regulated testing are different activities. A clinical laboratory sets its own cutoffs and panel for patient care, and chain of custody is not required. Federally regulated workplace testing uses fixed published cutoffs, a certified laboratory, a medical review officer, and full chain of custody. The Department of Health and Human Services (HHS) mandatory guidelines added fentanyl and norfentanyl to the federal urine panel at a 1 ng/mL cutoff effective July 7, 2025, and require a grouped initial immunoassay to cross-react with the other analytes in its group.^{1717 Substance Abuse and Mental Health Services…, et al. Mandatory guidelines for federal workplace drug testing programs. 2025.} Transportation-sector testing runs under a separate Department of Transportation regulation with its own procedural rules and its own cutoff table, which is not identical to the HHS table. The two tables come from different agencies, and each applies only to its own program.^{1818 Department of Transportation. 49 CFR part 40, Procedures for Transportation Workplace Drug and…} Documented chain of custody is required whenever results may be used for employment, legal, or forensic purposes.

Cross-reactivity can produce false-positive results, and poor antibody recognition can produce false-negative results.

Detection windows depend on the cutoff, assay, dose, use pattern, and individual metabolism. ARUP's estimated urine windows illustrate that variation: amphetamine 1 to 7 days, methamphetamine 1 to 4 days, MDMA 1 to 3 days, and benzoylecgonine 1 to 2 days. Benzodiazepine examples range from midazolam at 1 to 2 days to nordiazepam at 1 to 20 days. Codeine and morphine are estimated at 1 to 3 days, but 6-monoacetylmorphine (6-acetylmorphine) at under a day. Methadone is estimated at 1 to 14 days, buprenorphine at 1 to 10 days, phencyclidine at 1 to 7 days, and the urinary THC metabolite at 1 to 45 days. These are approximate analyte-specific windows. They cannot date an ingestion. Consult the performing laboratory's assay limits and interpret the use history.^{44 Mayo Clinic Laboratories. Test catalog: phenytoin, total and free, digoxin, serum… 2025.} Metabolic relationships complicate interpretation further: codeine is metabolized to morphine, morphine to a small amount of hydromorphone, hydrocodone to hydromorphone and norhydrocodone, oxycodone to oxymorphone, and poppy-seed ingestion can produce measurable urine morphine for several hours. Interpretation therefore considers the detected analytes together, since a single positive can have several explanations.

Specimen validity testing accompanies the federal drug panel because dilution, substitution, and adulteration are common. Creatinine is measured on every specimen, specific gravity when creatinine is low, along with pH and an oxidant test. A specimen with creatinine below 2 mg/dL and specific gravity at or below 1.0010, or at or above 1.0200, is substituted, since those values are not consistent with human urine. Intermediate low creatinine with low specific gravity is dilute. pH below 4 or at or above 11, or nitrite at or above 500 µg/mL, is adulterated. Invalid results have their own criteria, including specified inconsistent or interfering findings. Apply the program's criteria, with the required confirmation on separate aliquots, before assigning a classification. Report dilute findings with the positive or negative drug result. Adulterated, substituted, and invalid findings are reported in their own categories.^{1919 Substance Abuse and Mental Health Services…, et al. Mandatory guidelines for federal workplace drug testing programs… 2023.} Clinical laboratories may adopt some of these checks but are not bound by the federal criteria.

Athletic anti-doping work, including the urinary testosterone to epitestosterone ratio and isotope-ratio mass spectrometry, is performed by accredited specialist laboratories under separate rules and is outside routine clinical laboratory scope.

Genotype explains part of the variability that monitoring measures, and the two are complementary. A genotype predicts metabolic capacity before the first dose and never changes. A concentration reflects actual exposure, including the effects of adherence, interactions, and organ function. The relevant pairs among the drugs above are CYP2C9 with phenytoin, HLA-B*15:02 and HLA-A*31:01 with carbamazepine hypersensitivity risk, TPMT and NUDT15 with thiopurines, CYP2D6 with tricyclic antidepressants, and CYP3A5 with tacrolimus. FDA labeling recommends HLA-B*15:02 testing before starting carbamazepine in patients of at-risk ancestry, and TPMT and NUDT15 status informs thiopurine use. The Clinical Pharmacogenetics Implementation Consortium publishes the gene-drug guidelines that translate a genotype into a prescribing recommendation.^{2020 Clinical Pharmacogenetics Implementation… CPIC guidelines.} None of these results removes the need for concentration monitoring where a target range exists.