Blood Tests - Patient Investigations

Blood Tests

Why test

·        Before ordering any test always ask yourself why you are ordering it. Labs confirm a diagnosis – don‟t give it

·        Diagnosis: to confirm diagnosis/exclude differential diagnosis from history & exam

·        Prognosis: severity/progression

·        Monitoring: Measure target of treatment rather than drug level (e.g. INR rather than warfarin)

·        Screening: Only where test is reliable and you can do something about it

Parameters of a Test

·        Normal range: either arbitrary (level which leads to ¯risk) or statistical (what most people are)

·        Reliability:

o  Accuracy: mean of test results = real result 

o  Precision: variability in results (i.e. want a small standard deviation). Important for serial monitoring. Only different if 2½ SDs from previous test

·        Sensitivity: what rate of true positives does it pick up (are all positives found?)

·        Specificity: False negative rate

Test Results

·        Results may be:

o  Real & require interpretation

o  Erroneous: will always be some errors – there should be known rates of error for a lab and these should be within acceptable limits

·        Artefact: affected by non-disease factors:

o  Pre-analytical artefacts: mainly at time of collection

§  Incorrect labelling

§  Wrong tube/anticoagulant

§  Haemolysis

§  Delayed transport

§  Temperature effects e.g. refrigerating stuffs up electrolytes

§  Sample incorrectly taken (e.g. through or close to IV lines)

·        Pre-analytical factors:

o   Not fasted/wrong time for sample

o   Medications interfere

o   Wrong reference range

Urgent Tests

·        If the result may change the immediate management of a patient or if it plays a major role in on going assessment of a critically ill patient

·        Routine ordering/screening not appropriate in A&E

·        Emergency electrolytes:

o   Frequently over-ordered

o   Indications include D/V, seizure of unknown cause, muscle weakness, > 65, known renal/diabetes disease

·        Blood gases:

o   Don‟t need for uncomplicated asthma/MI, or if normal systemic perfusion and no dyspnoea/hyperventilation

o   Indicated if: cyanosis, severe dyspnoea, hypotension, vasoconstricted and sweaty, septic shock, pneumonia, suspected PE, CORD in acute exacerbation

·        Beware overdoses: people miscalculate/lie about consumption

o   Timing important: test for paracetamol overdose after 4 hours to judge treatment required. Changes in liver function take 24 hours

o   Ethanol levels: check in unconscious patient, for medicolegal reasons, or if intoxicated but potentially multiple problems

·        Toxicology Testing:

o   Serum levels for paracetamol, aspirin, ethanol, methanol, ethylene glycol, lithium, anticonvulscents, digoxin, iron, theophylline

o   Urine screen for drugs of abuse

o   Toxilab screen: long and slow for about 400 therapeutic drugs.  Qualitative only

o   Emergency use of cardiac markers: Beware timing - only after 6 hours unless as baseline. Can‟t size infarcts on cardiac enzymes

·        Abdominal pain:

o   Common to find no specific biochemical change

o   Baseline Na, K, creatinine if D/V or surgery likely

o   Amylase, glucose, HCG, LFT, calcium, cardiac enzymes

o   More rarely: urinary porphobilinogen, blood lead

o   Acute pancreatitis may not have ­amylase, and ­amylase can present in other conditions e.g. perforated/ischaemic bowel, ruptured ectopic pregnancy, diabetic ketoacidosis, renal failure