--- module: 064-01 language: en chapter: 64 title: "Common Laboratory Tests, Diagnostic Imaging, and Test Stewardship" module_title: "Foundations" source_sha256: 9e4cd9326d94359ec934e1d02558d5b20bd403170b0d734861027c9ff4dfc514 --- # Laboratory tests, imaging, and stewardship ## Orientation, probability and thresholds ### A test changes uncertainty, not reasoning #### Harms: pain, anaemia, radiation, cascades, delay #### Order when results change action ### State target condition and pretest probability ### Sensitivity and specificity #### Predictive values vary with prevalence #### Likelihood ratios convert pretest to post-test odds ### Sensitive negative excludes, specific positive confirms #### Fails if timing, technique, spectrum, cut-off differ ### Thresholds reflect consequences #### Dangerous treatable disease: act at lower probability ### Reference intervals hold 95% of healthy people #### Multiple testing adds chance abnormality ## Specimen quality and pre-analytical error ### Identity, timing, tube, site, transport ### Posture, fasting, medicines, tourniquet alter results ### Haemolysis raises potassium and enzymes ### Delay changes glucose and gases ### Underfilled citrate tubes distort coagulation ### Unexpected result: assess patient, check plausibility #### Repeat urgently, never dismiss critical results ### Closed loop: recipient, meaning, action, owner ## Full blood count and film ### A map rather than a diagnosis ### Mean cell volume normal with mixed deficiencies ### Reticulocytes adjusted for anaemia severity ### Neutropenia risk by depth and duration ### Platelet number does not measure function #### Bleeding risk: vessel, drugs, organs, procedure ### Film: fragments, blasts, parasites, clumping ## Electrolytes, kidney function and osmolality ### Sodium mainly reflects water balance #### Correct chronic change slowly ### Potassium: sampling, acid-base, insulin, drugs #### Severe or electrical change: treat immediately ### Creatinine imperfect and lags acute injury #### Estimated filtration unreliable when changing ### Urinalysis shows patterns serum cannot ### Osmolar and anion gaps #### Neither excludes dangerous poisoning ## Liver, cardiac, inflammatory and endocrine markers ### Aminotransferases indicate hepatocyte injury #### High enzymes do not mean worse function ### Troponin indicates injury, not infarction #### Infarction: rise or fall plus ischaemia ### Natriuretic peptides reflect wall stress ### CRP and ESR nonspecific inflammation ### Autoantibodies need compatible phenotype ### Treat suspected adrenal crisis before testing ### Glycated haemoglobin distorted by red-cell change ## Coagulation and transfusion ### Prothrombin time for vitamin-K antagonists ### Normal tests miss platelet and von Willebrand defects ### Abnormal tests: deficiency, inhibitor, error ### Group, screen and crossmatch prevent incompatibility #### Uncrossmatched blood when haemorrhage outweighs delay ### Transfuse by physiology, not one threshold ## Microbiology and molecular testing ### Sample infected site before antimicrobials ### Positive culture: pathogen, coloniser, contaminant ### Susceptibility does not ensure penetration ### PCR may stay positive after organisms die ### Serology may be negative early ## Plain imaging, ultrasound, CT and MR ### Radiography: overlap hides disease #### Normal chest film misses early infection, small pneumothorax ### Ultrasound: no radiation, operator dependent ### CT: rapid cross-sectional anatomy #### Radiation accumulates, but never delay life-saving scans #### Iodinated contrast: weigh benefit and kidney context ### MR: soft-tissue contrast, slower, implant limits #### Gadolinium cautiously in kidney failure or pregnancy ## Nuclear and interventional imaging ### Radiotracers map physiology #### Inflammation mimics cancer ### Interventional: biopsy, drainage, embolisation #### Biopsy route must preserve later surgery ## Stewardship and incidental findings ### Ask question, result impact, timing, harm ### Stop monitoring that no longer guides care ### Least harmful test that answers the question ### Screening: low prevalence, overdiagnosis ### Describe incidental findings by risk and follow-up ### Assign ownership at every transition