--- module: 064-02 language: en chapter: 64 title: "Common Laboratory Tests, Diagnostic Imaging, and Test Stewardship" module_title: "Measurement uncertainty, longitudinal interpretation, imaging choice, and result ownership" source_sha256: d7158592e90bc1b94d0d072360cd4136cb33576bf43f5fbd60004819b71e89d0 --- # Measurement uncertainty and result ownership ## Tests as probability interventions ### Value only when result changes care ### Harms: false reassurance, false alarms, cost, delay ### Before ordering: target, pretest, threshold, action ### Sensitivity and specificity tied to population and threshold ### Predictive values shift with prevalence #### Low-risk positives mostly false ### Published performance may not transport ## Thresholds and continuous values ### Testing threshold between treat and exclude #### Neither crossed: repeat, new principle, observe ### Harmful false negative lowers exclusion threshold ### Invasive treatment harm raises treatment threshold ### Values, follow-up, disease speed move thresholds ### Continuous values beat binary flags #### Potassium 6.8 is not 5.2 ### One in twenty healthy results outside range #### Twenty tests make a flag likely ## Pre-analytical error and timing ### Occurs before the analyser sees the specimen ### Haemolysis releases potassium and enzymes ### Underfilled citrate: excess anticoagulant, long clotting ### Blood gas changes with air and metabolism ### Critical values: assess patient, not auto-dismiss ### Sample time recorded relative to treatment ### Line sampling: discard volume, stop infusion ### Scarce fluids: prioritise tests in advance ## Post-analytical error and ownership ### Units, ranges, transcription, inbox routing fail ### Unfinished until a named clinician acts #### Electronic delivery alone is not closed loop ### Pending results at discharge need owner ### Critical call: read back identity, value, time #### Use escalation route if clinician absent ### Tell patients: confirmed or preliminary, urgency ### Ownership and named alternate prevent misses ## Blood count and electrolytes ### Normal cell volume with iron plus B-twelve deficiency ### Reticulocytes corrected for severity and maturation ### Neutropenia risk: count, duration, barrier, immunity ### Sodium: water relative to exchangeable cations #### Acute severe low sodium: cerebral oedema #### Rapid chronic correction: osmotic demyelination ### Pseudohyperkalaemia from haemolysis, fist clenching #### Treat electrical instability while confirming #### Normal tracing does not make severe levels safe ### Magnesium deficiency: refractory potassium ## Kidney, liver, cardiac and endocrine markers ### Creatinine delayed and imperfect #### Estimated filtration assumes steady state ### Urine studies explain mechanism better ### Liver: injury, cholestasis, function #### Falling enzymes in failure: loss of viable cells ### Troponin proves myocardial injury #### Infarction needs rise or fall plus ischaemia ### Natriuretic peptides lower in obesity ### Endocrine tests measure feedback systems #### Treat suspected adrenal crisis before testing ## Coagulation and microbiology ### Screens sample limited pathways ### Mixing study: deficiency versus inhibitor ### Bleeding history outweighs indiscriminate screening ### Deep specimen before antibiotics establishes cause ### Superficial swabs report colonisers ### Nucleic-acid tests detect dead organisms and carriage ### In vitro susceptibility not penetration ## Imaging modalities and contrast ### Radiography: projection with overlap ### Point-of-care ultrasound is not a formal study ### CT: rapid anatomy, cumulative radiation ### Iodinated contrast: kidney risk depends on context #### Unrelated allergy does not prohibit contrast #### Premedication does not eliminate reactions ### Delay for theoretical renal risk can harm more ### Magnetic resonance not best just for no radiation ## Imaging choice and incidental findings ### Non-contrast CT for acute intracranial blood ### Diffusion imaging for early infarction ### Ultrasound first for gallstones ### CT angiography for mesenteric ischaemia ### Least harmful test that adequately answers ### Incidental findings: second diagnostic problem #### Avoid catastrophic language, never bury action ### Overdiagnosis: screening needs stronger evidence ## Longitudinal stewardship ### Stop daily panels that no longer guide care ### Repeat when trajectory is the signal ### Document reason, probability change, action, owner ### Record decisions not to test