CFB MED10 The Use Of Laboratory Test In Clinical Medicine
The systematic application of laboratory investigations—including selection, interpretation, sensitivity, specificity, and predictive values—to aid in disease screening, diagnosis, monitoring, and prognosis in clinical practice.
The Use of Laboratory Tests in Clinical Medicine
Lecture Map
This lecture is about thinking before you order, and thinking again when you read the result. It covers the entire journey of a laboratory test — from the clinical question that prompts the order, through pre-analytical pitfalls, analytical interferences, to post-analytical interpretation traps. The overarching philosophy is "Primum non nocere" — even lab tests can harm if misused, misinterpreted, or ordered without purpose. [1]
- Understand the Laboratory Testing Cycle and its phases.
- Know the reasons for ordering laboratory tests (confirm, aid, refine, monitor, assess, detect, stratify, optimize).
- Appreciate reference intervals and their statistical basis — why "out of range" ≠ disease.
- Recognise pre-analytical, analytical, and post-analytical pitfalls that produce spurious results.
- Understand specific interference mechanisms: haemolysis, in vitro glycolysis, high-dose hook effect, biotin interference, heterophilic antibodies, macro-complexes, drug interferences.
- Understand the role and caveats of Point-of-Care Testing (POCT).
- Apply precision medicine concepts: pharmacogenomic testing (HLA alleles, DPD deficiency, molecular markers for targeted therapy).
The laboratory testing cycle has three phases: pre-analytical, analytical, and post-analytical. Most errors (~60-70% in real-world data) occur in the pre-analytical phase. [1]
Why this matters: Clinicians control the pre-analytical phase (ordering the right test, correct specimen, correct tube, correct timing, correct labelling). They also control the post-analytical phase (interpretation). The analytical phase is largely the lab's domain but clinicians must still understand interferences.
"Primum non nocere" — Seven questions to ask before ordering: [1]
- Why do I request this test?
- What are the consequences of NOT performing the test?
- How good is the test at discriminating health from disease? (sensitivity/specificity)
- What will I look for in the result?
- How are results interpreted?
- How will results influence management and outcome?
- Will this ultimately benefit the patient?
Why this matters clinically: Unnecessary tests generate false positives (recall: 5% of healthy people fall outside the reference interval per test), cause patient anxiety, lead to invasive follow-up procedures, and waste resources. Every test must have a clinical question attached to it.
Principles of Cooperation with the Laboratory (Hutchison's Clinical Methods)
Clinicians should know: [1]
- What investigation to request
- The discomfort and possible risk to the patient of an investigation
- What information the pathologist requires about the patient
- What specimens are required, and how they should be obtained and transmitted to the laboratory
- The approximate cost, in time and money, of an investigation
- The possible risk of a specimen to the laboratory worker (e.g., infected material)
- How to make the best use of the results received from the laboratory
- The investigation plan should be decided from facts elicited by history-taking and physical examination — the availability of routine tests should not lead to abandonment of the elective approach to investigation.
- Monitoring investigations should follow a planned series of repeat tests at logically decided intervals — tests should not be repeated without good reason. [1]
| Purpose | Example | Why It Matters |
|---|---|---|
| CONFIRM a diagnosis | Fasting glucose or OGTT for DM; TSH & fT4 for thyroid dysfunction; troponins for AMI; toxicology screen in acute confusion; plasma amino acid & urine organic acid profiles in IEM | Converts clinical suspicion into a definitive diagnosis |
| AID differential diagnosis | LFT, viral serology, autoimmune markers to differentiate hepatitis types | Narrows the diagnostic list systematically |
| REFINE a diagnosis | ACTH to differentiate ACTH-dependent vs ACTH-independent Cushing; central vs nephrogenic DI | Determines the specific subtype to guide management |
| MONITOR progress | Serial AFP in HCC; fT4 in thyrotoxicosis on treatment; HbA1c in DM; IGF-1 in GH supplementation | Tracks whether disease is improving, stable, or worsening |
| ASSESS severity | Creatinine & eGFR in renal impairment; CBP in anaemia; coagulation profile in bleeding diathesis; ketones/pH in DKA; CK in rhabdomyolysis | Guides intensity of treatment (e.g., DKA with pH < 7.0 → ICU) |
| DETECT complications / side effects | CBP during cytotoxic therapy; RFT during nephrotoxic therapy; K⁺ in patients on diuretics | Early detection prevents irreversible harm |
| MONITOR therapy (TDM) | Anti-epileptics, immunosuppressants, aminoglycosides, theophylline, lithium | Drugs with narrow therapeutic windows need level monitoring to avoid toxicity or subtherapeutic dosing |
| STRATIFY risk | Full lipid profile for cardiovascular risk stratification | Prevention is better than cure |
| OPTIMIZE outcome / MINIMIZE adverse effects | DPYD genotype for fluoropyrimidine therapy; HLA B*15:02 for carbamazepine | Precision medicine: right drug, right dose, right patient |
Era of "personalised" or "precision medicine": Pharmacogenetic and genomic testing allows better prediction of adverse drug effects or therapeutic efficacy [1]
| Genetic Marker | Drug | Clinical Consequence |
|---|---|---|
| HLA B15:02* | Carbamazepine | Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) — Must screen before prescribing in at-risk populations (Southeast Asian, including Hong Kong Chinese) |
| HLA B58:01* | Allopurinol | Severe cutaneous adverse reactions (SCAR) — Screen before prescribing |
| DPYD deficiency | 5-FU, capecitabine, tegafur | Enhanced toxicity due to reduced drug clearance (dihydropyrimidine dehydrogenase is the rate-limiting enzyme for fluoropyrimidine catabolism) |
| Molecular markers | Targeted cancer therapy | E.g., HER2 in breast cancer, EGFR in lung cancer, BRAF in colorectal cancer |
Why HLA B*15:02 matters for HKU students: This is high prevalence in Southern Chinese (HK population ~8% carrier rate). It is now standard practice to test before prescribing carbamazepine. If positive → absolute contraindication. This is a commonly examined pharmacogenomics example. [1][8]
Key considerations for population screening: [1]
- The condition is common or life-threatening
- The tests are readily applicable and acceptable to the population
- The tests are sensitive and specific
- Facilities exist for follow-up and confirmation
- Economic impact has been assessed and accepted
Case-Finding Programme Examples
| Population | Condition | Test |
|---|---|---|
| Neonates | Congenital hypothyroidism, G6PD deficiency | Cord blood TSH ± fT4, G6PD |
| Neonates | IEM (26 conditions in HK) | Dried blood spots for amino acids, acylcarnitines (tandem mass spectrometry) |
| Adolescents/Young Adults | Substance abuse | Urine drug screen |
| Pregnancy | GDM | Plasma glucose, OGTT |
| Pregnancy | Open neural tube defect | Maternal serum α-fetoprotein |
| Industry | Lead exposure | Whole blood lead levels |
| Industry | Organophosphate exposure | Serum pseudocholinesterase activity |
| Elderly | Malnutrition | Vitamin D, prealbumin, retinol binding protein |
| Elderly | Thyroid dysfunction | TSH ± fT4 |
High Yield – Newborn Screening in HK
Hong Kong currently screens for 26 inborn errors of metabolism (IEM) conditions using dried blood spots for amino acids and acylcarnitines, in addition to congenital hypothyroidism and G6PD deficiency. This was expanded following an 18-month pilot study. [1]
The 26 IEM conditions comprise:
| Category | Number | Examples |
|---|---|---|
| Disorders of organic acids | 8 | Methylmalonic acidaemia, propionic acidaemia, glutaric acidaemia type I, isovaleric acidaemia, multiple carboxylase deficiency, beta-ketothiolase deficiency, 3-hydroxy-3-methylglutaryl-CoA lyase deficiency, MMA with homocystinaemia (cobalamin C deficiency) |
| Disorders of amino acids | 9 | Classical PKU, 6-pyruvoyl-tetrahydropterin synthase deficiency, maple syrup urine disease, homocystinuria, citrullinaemia types I & II, argininosuccinic acidaemia, argininaemia, tyrosinaemia type I |
| Disorders of fatty acid oxidation | 6 | MCADD, VLCADD, glutaric acidaemia type II, carnitine uptake deficiency, carnitine-acylcarnitine translocase deficiency, CPT II deficiency |
| Other IEM | 3 | Congenital adrenal hyperplasia, biotinidase deficiency, classic galactosaemia |
In the 18-month pilot study, 9 babies were confirmed positive — 8 were asymptomatic at the time of screening (only 1 baby with methylmalonic acidaemia was symptomatic). The most common diagnosis was carnitine uptake deficiency (3 cases). [1]
6. Reference Intervals ("Normal Ranges")
Reference intervals are based on the central 95th percentile (Mean ± 2SD for Gaussian data, or 2.5th–97.5th centile for non-parametric data) of a representative sample of healthy individuals. [1]
- ~5% of normal subjects will have results outside the reference interval — this is by definition, not disease
- Deviation from the reference interval does NOT necessarily mean disease; the converse is also true (i.e., a "normal" result does not exclude disease)
- Reference intervals vary by age, gender, race, and test methodology → values from literature may not be locally applicable
- Always interpret in the proper clinical context
Why this is exam-critical: A common MCQ trap is presenting a result slightly outside the reference interval and asking whether the patient has disease. The answer depends on clinical context, not the number alone.
Rule of thumb: If results from specimens collected on different occasions differ by more than 2.77 × total SD (analytical + biological variation combined), there is > 95% chance of a genuine change. [1]
Why this matters: When monitoring a patient (e.g., serial creatinine in AKI), small fluctuations may be within normal analytical/biological variation. Only changes exceeding this threshold should prompt clinical action.
7. Pitfalls and Artefacts — Pre-Analytical Factors
High Yield – Artefactual Results
These pre-analytical errors are extremely commonly tested in MCQs. Memorize the direction of change for each. [1]
| Factor | Effect on Lab Results | Mechanism |
|---|---|---|
| Prolonged venous stasis | ↑ Total Calcium, Albumin, Lipids | Fluid shifts out of the intravascular space → haemoconcentration of large molecules |
| Stress | ↑ Prolactin, ACTH, Cortisol | Hypothalamic-pituitary axis activation |
| Smoking | ↑ Ammonia | Tobacco smoke contains ammonia |
| Excess heparin | ↓ Ionised calcium, electrolytes | Heparin chelates calcium; dilution effect |
| EDTA contamination | ↓ Ca, ALP, Mg; ↑ K | EDTA chelates divalent cations (Ca²⁺, Mg²⁺) and is a potassium salt (K₂-EDTA or K₃-EDTA) |
| Alcohol swab | ↑ Ethanol (↑ isopropanol) | Contamination of specimen if skin not dry before venepuncture |
| Li-heparin tubes | ↑ Lithium level | Tube anticoagulant is lithium heparin → spurious therapeutic lithium level |
| Collection from drip-arm | Dilution/contamination | IV fluid dilutes the specimen or adds glucose/electrolytes from the infusate |
| Acid-washed tubes | ↑ Aluminium, Zinc, Manganese, Selenium | Trace metals leach from inadequately washed containers |
A 46-year-old woman on steroids for RA, blood sample taken at the Rheumatology Clinic on a Saturday morning: Na 128, K 6.7, Cl 89, Urea 5.0, Creatinine 105, Glucose 2.3. [1]
Looks like Addisonian crisis (low Na, high K, low glucose). BUT — the clue is "Saturday morning at the Rheumatology Clinic." This was likely an EDTA-contaminated specimen (e.g., blood drawn into EDTA tube then decanted into a biochemistry tube, or cross-contamination). EDTA causes spuriously low Ca and high K. The low glucose is from in vitro glycolysis due to delayed processing over the weekend.
How to spot this in exams: Whenever Na is low, K is very high, and Cl is low but urea/creatinine are normal → think EDTA contamination + delayed specimen processing, not genuine electrolyte disorder.
In Vitro Changes in Blood Specimens
After blood collection, erythrocytes continue to metabolize glucose (glycolysis), causing: [1]
- Glucose ↓ (up to 10% in the first hour)
- K⁺ ↑ (K⁺ leaks out of cells)
- Na⁺ shifts into cells
- HCO₃⁻ ↓ (moves out), CO₂ ↑ (moves out), Cl⁻ shifts into cells
The rate of in vitro glucose decrease is accelerated by increased ambient temperature and increased cell counts (e.g., leukaemia with very high WBC). [1]
Mitigation:
- Fluoride-oxalate tube inhibits glycolytic enzymes (but takes ~1 hour to reach maximum inhibition — so it's not instantaneous!)
- Prompt separation of plasma from cells is the KEY — the most reliable way to prevent glycolysis-related glucose drop
- Gel separator tubes after centrifugation create a physical barrier between plasma and cells
Why this matters: If you send a glucose specimen to a distant lab without fluoride or without rapid processing, the glucose value will be falsely low, potentially leading to misdiagnosis of hypoglycaemia or underestimation of hyperglycaemia in a diabetic.
Haemolysis → falsely LOW insulin (insulin is degraded by erythrocyte insulinase released during haemolysis) [1]
Insulin-binding antibodies → falsely HIGH insulin (antibodies bind insulin and prevent clearance, or interfere with the assay depending on the method; some methods show falsely low results)
Clinical pearl: If insulin results don't match the clinical picture (e.g., patient supposedly hypoglycaemic with "low" insulin but you suspect insulinoma), consider C-peptide as an alternative marker. C-peptide is co-secreted with insulin in equimolar amounts but is not affected by haemolysis or insulin antibodies. This is why C-peptide is preferred in the workup of hypoglycaemia. [1]
10. High-Dose Hook Effect
In "one-stage 2-site sandwich immunometric assays," gross antigen excess can saturate both the capture and detection antibodies independently (rather than forming the intended sandwich), resulting in a falsely LOW or NORMAL measured value despite extremely high actual concentration. [1]
| Zone | What Happens | Result |
|---|---|---|
| Antibody excess | Normal assay function | Accurate |
| Zone of equivalence | Optimal Ag-Ab complex formation | Accurate (peak signal) |
| Antigen excess | Both Ab saturated independently → no sandwich formed | Falsely LOW |
AFP, Ferritin, Prolactin, β-hCG, and other tumour markers [1]
55-year-old man, AFP was 24.0 ng/mL pre-operatively, then "rose" to 136 ng/mL 2 weeks after left hepatic lobectomy for HCC. Paradoxical? Actually, the pre-operative AFP was falsely low due to hook effect. The true pre-operative AFP was astronomically high, and the post-operative 136 was the declining (but still elevated) true value after tumour debulking. [1]
Exam trap: When a tumour marker is "unexpectedly low" in the setting of known high-volume malignancy, always consider the hook effect. Request a diluted specimen to unmask the true value.
Hook Effect — Clinical Rule
If a tumour marker result is paradoxically low or normal in a patient with high clinical suspicion of advanced malignancy, request serial dilutions of the specimen. The hook effect causes falsely low results in antigen excess. [1]
11. Biotin Interference
Many modern immunoassays use the biotin-streptavidin interaction as part of their chemistry (biotin binds streptavidin with extremely high affinity, used to anchor reagents). Exogenous biotin in the patient's blood competes with the assay's biotin-labelled reagents for streptavidin binding sites.
| Assay Type | Example | Direction of Error |
|---|---|---|
| Immunometric (sandwich) assay | TSH | Falsely LOW |
| Competitive assay | Free T4 | Falsely HIGH |
Why opposite directions?
- In sandwich assays: Biotin displaces the biotin-labelled detection antibody from the streptavidin-coated surface → less signal captured → falsely low reading
- In competitive assays: Biotin displaces the biotin-labelled analyte analogue from streptavidin → less analogue bound → system interprets this as more patient analyte present → falsely high reading
A patient taking high-dose biotin supplements (common in hair/nail health products, or therapeutic doses in multiple sclerosis) could present with:
- Low TSH + High fT4 → mimicking thyrotoxicosis when the patient is actually euthyroid
- This has led to inappropriate thyroid treatment in real cases
Rapid screening kits are commercially available (e.g., VeraTest Biotin®) — serum biotin > 15 ng/mL gives a positive screen result [1]
Management: Ask patients to stop biotin supplements for at least 48-72 hours before blood tests.
The lecture highlights two landmark reports:
- Misdiagnosis of Graves' disease — A patient with progressive multiple sclerosis on megadose biotin presented with extremely elevated fT4/T3, suppressed TSH, and elevated TSH receptor antibodies. The patient had no symptoms of hyperthyroidism. Discontinuation of biotin for ≥2 days resulted in complete resolution of all biochemical abnormalities. [1]
- Paediatric series — Six children receiving high-dose biotin (2–15 mg/kg/day) for inherited metabolic diseases all had laboratory results mimicking Graves' disease; antithyroid medication was inappropriately initiated in at least three. [1]
Questioning patients about biotin intake should be part of the evaluation of any patient undergoing endocrine immunoassay testing. [1]
12. Other Analytical Interferences
| Interferent | Mechanism | Example |
|---|---|---|
| Lipids, haemoglobin, bilirubin | Spectrophotometric interference (turbidity, absorbance overlap) | Lipaemic serum → falsely elevated bilirubin or low sodium |
| Heterophilic antibodies / HAMA | Cross-link assay antibodies without analyte present → false signal | False positive tumour markers after mouse-derived antibody exposure |
| Anti-analyte antibodies | Anti-thyroglobulin Ab → low thyroglobulin in thyroid cancer monitoring; Anti-insulin Ab → spurious insulin levels | Thyroid cancer follow-up with unreliable Tg |
| Macro-complexes | Analyte bound to immunoglobulin → delayed clearance, not bioactive | Macro-prolactin (falsely elevated prolactin but no clinical hyperprolactinaemia), macro-TSH, macro-amylase, macro-CK, macro-transaminase |
| Paraproteins | Interfere with multiple assay platforms | Spurious electrolyte results, pseudohyponatraemia |
| Eltrombopag | Spectrophotometric interference on total bilirubin assay (orange-coloured plasma) | Falsely elevated bilirubin |
| Teriflunomide (active metabolite of leflunomide) | Falsely decreased ionised calcium | Analytical interference on ion-selective electrode |
Suspect spurious results when: [1]
- Result is clinically unexpected and inconsistent with other correlates
- Inconsistent with other parameters in multi-parameter testing
- Significant unexpected change from previous test
- Grossly and permanently "abnormal" in an apparently healthy subject
- Increased likelihood of interference: recent immunisation, blood transfusion, immunotherapy, imaging with fluorescent compounds, prolonged contact with animals (heterophilic antibodies)
The following drugs raise serum creatinine without actually impairing GFR: [1]
| Drug | Mechanism |
|---|---|
| Cimetidine | Inhibits tubular secretion of creatinine |
| Septrin (Co-trimoxazole = sulfamethoxazole/trimethoprim) | Trimethoprim inhibits tubular secretion of creatinine |
| Fenofibrate | Increases creatinine production |
| Cephalosporins (IV) | Analytical interference (Jaffé method) |
| Creatine supplements | Precursor to creatinine |
| Prednisolone | Increases muscle catabolism |
| Pyrimethamine | Inhibits tubular secretion |
Why this matters: Do not confuse a drug-induced creatinine rise with worsening renal function. If in doubt, measure cystatin C (not affected by tubular secretion) or calculate creatinine clearance from a timed urine collection. [1][2]
14. Point-of-Care Testing (POCT)
POCT is clinical laboratory testing conducted close to the site of patient care, providing rapid turnaround of test results for prompt patient management. [1]
Why POCT exists: In conventional laboratory testing, specimen delivery to the lab constitutes a significant proportion of the total turnaround time (TAT). For time-critical decisions (e.g., DKA management, blood gas in respiratory failure, glucose in hypoglycaemia), waiting for central lab results could be dangerous.
Specimen delivery to the lab makes up a significant proportion of the total test TAT. For some cases, specimen integrity may also be jeopardized during delivery (e.g., ammonia, blood glucose). [1]
Many POCT devices (urine drug screens, rapid antigen tests) use lateral flow assay technology:
- Sample applied to the sample pad
- Flows across the conjugate pad (contains labelled antibodies)
- Migrates along the nitrocellulose membrane past the test line (T) and control line (C)
- Excess fluid absorbed by the absorbent pad
Reading results: A visible control line (C) confirms the test is valid. A visible test line (T) indicates a negative result (for competitive format) or positive result (for sandwich format), depending on the assay design. If no control line appears, the test is invalid. [1]
| Application | Device/Test | TAT |
|---|---|---|
| Blood gases and electrolytes | Bedside blood gas analyser | ~4 minutes |
| Blood glucose | Glucometer | ~5 seconds |
| Urine drug screen | Lateral flow immunoassay | ~5 minutes |
| Cardiac markers | Point-of-care troponin | ~15 minutes |
| Infectious disease | Rapid antigen test (e.g., COVID, influenza, strep) | ~15 minutes |
| Disease panels | Piccolo Xpress (metabolic panel) | ~12 minutes |
| Ammonia | Bedside NH₃ (analyte prone to in vitro changes) | Minutes |
| Nucleic acid amplification | GeneXpert (TB, flu, COVID) | ~1 hour |
Errors in POCT diagnostics can lead to misdiagnosis, improper treatment, costly follow-up, death, and medico-legal consequences. [1]
| Pitfall | Explanation |
|---|---|
| Environmental exposure | Reagents exposed to humidity, temperature fluctuations → degradation |
| Pre-analytical errors | Patient ID errors, specimen collection/handling errors; most interfering factors occur pre-analytically |
| Operator experience | Pre-analytical errors have an inverse association with operator experience → adequate training is critical |
| Bubbles, micro-clots, gross clotting | Occur during specimen transfer/loading if procedure not followed |
| Biotin interference | Can affect HIV POCT and some tumour markers |
| GDH-PQQ glucose monitoring | Reports erroneously elevated glucose in patients exposed to maltose, icodextrin (peritoneal dialysis fluid), maltodextrin, galactose, or xylose → can mask life-threatening hypoglycaemia |
| High haematocrit | Some glucose meters inaccurate in neonates (high Hct) |
| Haemolysis detection | Cannot detect haemolysis in whole blood/fingerstick specimens → spurious K⁺ elevation goes undetected |
| Compromised peripheral circulation | Sepsis, shock, DKA → inadequate capillary blood samples → inaccurate results |
| Urine drug screen false positives | Not definitive; prone to false positives and false negatives; requires GC-MS/MS or LC-MS/MS confirmation |
POCT Urine Drug Screen
POCT urine drug screens are immunoassay-based and are NOT definitive. They are prone to false positives (cross-reactivity with medications) and false negatives. Always confirm with GC-MS/MS or LC-MS/MS, especially when results have legal consequences. [1]
15. Case Illustrations from the Lecture
Context: CF patients require sweat chloride testing (sweat test). POCT conductivity-based sweat analyzers can give misleading results if sweat volume is insufficient or if the skin is contaminated. Always ensure adequate sweat stimulation (pilocarpine iontophoresis) and volume.
Context: Pre-eclampsia workup includes urine protein (spot protein:creatinine ratio or 24-hour urine), LFT, CBP/platelets, LDH, urate. Be aware that haemolysis in the specimen may affect results and that pregnancy itself alters reference intervals (e.g., ALP is physiologically elevated from placental isoenzyme, creatinine is lower due to increased GFR).
The catch in this case: The slide data shows the haemolysis index jumped from 9 to 351 between two specimens collected on the same day, while the urine protein/creatinine ratio rose dramatically from 26 to 1677 mg/mmol Cr. A grossly haemolysed specimen can interfere with multiple analytes — always check specimen quality indices before interpreting results in a high-stakes clinical scenario. [1]
Context: In neonatal jaundice, the key test is total and direct bilirubin. Be aware that POCT transcutaneous bilirubinometry may be inaccurate in certain skin tones or after phototherapy. Always confirm with laboratory serum bilirubin if values are near exchange transfusion thresholds.
Context: After IV iron (isomaltoside) administration, serum ferritin rises dramatically (iron is being stored). A dramatically elevated ferritin post-IV-iron does NOT indicate iron overload — it reflects the administered dose. Wait at least 7-14 days after IV iron before rechecking iron studies.
Context: A 34-year-old woman with AML started on histone deacetylase (HDAC) inhibitors. Serial biochemistry shows urate plummeting from 283 to < 30 µmol/L over 5 days with a rising LDH (153 → 249). The extremely low urate is due to rasburicase (recombinant urate oxidase) given for tumour lysis syndrome (TLS) prophylaxis/treatment — it enzymatically degrades uric acid. The rising LDH reflects ongoing tumour cell lysis. Do not interpret the low urate as reassuring in isolation — it is a pharmacological effect, not a sign of low tumour burden. [1]
Context: HbA1c is unreliable in the presence of anaemia, particularly if there is active/recent blood loss, haemolysis, or recent transfusion. This is because HbA1c reflects average glucose over the RBC lifespan (~120 days). If RBC lifespan is shortened (haemolysis, bleeding) or if new non-glycated RBCs flood the circulation (transfusion, iron replacement after bleeding), HbA1c will be falsely low. Use fructosamine or glycated albumin as alternatives. [1]
Context: In patients with heavy proteinuria, proteins lost in urine include albumin → low serum albumin → altered total calcium (but ionised calcium may be normal). Also, in progressing nephropathy, creatinine-based eGFR becomes less reliable and cystatin C-based equations may be preferred. Additionally, the HbA1c may be affected by uraemia (carbamylated haemoglobin interferes with some assays).
The catch in this case: The slide data shows an exceedingly high urine creatinine concentration (42,731 µmol/L) with a urine volume of only 0.32 L over 24 hours. The laboratory comment reads: "Suspected incomplete urine collection and/or significant renal failure." This illustrates that a spot urine protein:creatinine ratio can be misleading if the urine collection is incomplete or the specimen is highly concentrated — always assess urine volume and collection adequacy before interpreting 24-hour urine results. [1]
While not a dedicated slide topic, this concept underpins the entire lecture:
| Property | Definition | Clinical Use |
|---|---|---|
| Sensitivity | Proportion of truly diseased people who test positive (TP / [TP + FN]) | High sensitivity → good for ruling OUT disease (SnNOut) |
| Specificity | Proportion of truly non-diseased people who test negative (TN / [TN + FP]) | High specificity → good for ruling IN disease (SpPIn) |
Application from 2025 MCQ Q6
A lab test for a rare inherited metabolic disorder has very high specificity but moderate sensitivity. Which scenario BEST suits this test? [7]
Answer: A. Confirmatory testing after a positive screening test.
Reasoning: For screening, you want high sensitivity (don't miss cases). For confirmation, you want high specificity (don't label healthy people as diseased). A test with high specificity is ideal for confirming a positive screening result.
17. Integration with Related Material
- When investigating anaemia, CBC parameters (Hb, MCV, MCH) are the starting point; reticulocyte count helps distinguish decreased production from increased destruction/loss [3]
- Iron studies (serum iron, TIBC, ferritin) are interpreted together — TIBC is the most useful test to distinguish iron deficiency from anaemia of chronic disease [3]
- Reinforces the POCT concepts covered in this lecture
- Emphasises governance: POCT must have quality assurance, regular calibration, and operator training [1]
18. Exam Intelligence
| Year | Question | Topic | Key Point |
|---|---|---|---|
| 2021 Q7 | Assay interference giving spurious results | Prolonged venous stasis → ↑Ca, Albumin, Lipids (Answer: C) | Other options were correct practices, not sources of error [6] |
| 2025 Q6 | High specificity, moderate sensitivity | Best for confirmatory testing after positive screen | Understand SnNOut vs SpPIn [7] |
- "Normal" result excludes disease → FALSE. 5% of normal people fall outside reference intervals; conversely, early disease may have "normal" results.
- EDTA tube used for chemistry → Gives falsely low Ca²⁺, Mg²⁺, ALP and falsely high K⁺. A very common exam scenario.
- Drip-arm blood → Dilution artefact. Results are meaningless. Always collect from the opposite arm.
- Hook effect → Low/normal tumour marker in advanced cancer = hook effect. Request serial dilutions.
- Biotin → Low TSH + High fT4 ≠ thyrotoxicosis. Ask about supplements.
- HbA1c in anaemia/haemolysis/transfusion → Unreliable. Use fructosamine or glycated albumin.
- POCT glucose in peritoneal dialysis patients → GDH-PQQ meters cross-react with icodextrin → falsely elevated glucose → masked hypoglycaemia.
- Stress-induced prolactin elevation → Not a prolactinoma. Repeat after rest, or measure macroprolactin.
When asked "What are the possible causes of this spurious result?", structure your answer:
- State the phase of the lab testing cycle where the error occurred (pre-analytical > analytical > post-analytical)
- Name the specific factor (e.g., EDTA contamination, haemolysis, delayed processing)
- Explain the mechanism (e.g., EDTA chelates divalent cations; K₂-EDTA releases K⁺ into specimen)
- State the expected direction of change (e.g., ↓Ca, ↓Mg, ↑K)
High Yield Summary
-
The Lab Testing Cycle
- has pre-analytical, analytical, and post-analytical phases. Most errors are pre-analytical.
-
Seven questions
- must be answered before ordering any test — the principle of "Primum non nocere" applies.
-
Eight purposes
- of lab testing: Confirm, Aid differential, Refine, Monitor progress, Assess severity, Detect complications, Monitor therapy (TDM), Stratify risk, Optimize outcome (pharmacogenomics).
-
Reference intervals
- cover the central 95% of healthy individuals — ~5% of normal people fall outside by definition. Out-of-range ≠ disease; in-range ≠ health.
-
Pre-analytical pitfalls
- EDTA contamination (↓Ca, ↑K), prolonged venous stasis (↑Ca, Alb), drip-arm collection, in vitro glycolysis (↓glucose, ↑K), haemolysis, stress, smoking.
-
Hook effect
- Falsely low tumour markers in antigen excess — AFP, ferritin, prolactin, β-hCG. Request dilutions.
-
Biotin interference
- Falsely LOW sandwich assays (TSH), falsely HIGH competitive assays (fT4). Stop biotin 48-72h before testing.
-
POCT
- Rapid TAT but beware of GDH-PQQ glucose meters with icodextrin, haemolysis detection in whole blood, urine drug screen false positives (confirm with MS).
-
Drugs raising creatinine without affecting GFR
- Cimetidine, trimethoprim, fenofibrate, cephalosporins (Jaffé), creatine.
-
Pharmacogenomics
- HLA B15:02 (carbamazepine → SJS/TEN), HLA B58:01 (allopurinol → SCAR), DPYD (fluoropyrimidines → enhanced toxicity).
Active Recall - Laboratory Tests in Clinical Medicine
[1] Lecture slides: CFB (MED10) The use of laboratory test in clinical medicine.pdf [2] Senior notes: Block A – Nephrology Data Interpretation.pdf [3] Senior notes: Block A - Pallor_ diagnosis of anaemia; nutritional anaemia; anaemia of systemic diseases.pdf [4] Senior notes: Adrian Lui Pediatrics Notes.pdf (Section 10.2.3 Coagulation Disorders) [5] Senior notes: Block A - Gastrointestinal Data Interpretation.pdf [6] Past papers: 2021 Fourth Summative Assessment MCQ.pdf (Q7) [7] Past papers: 2025 Fourth Summative MCQ.pdf (Q6) [8] Lecture slides: GC 070. Is this the best drug for me.pdf