Paediatric Investigations and Diagnosis: Choosing Relevant Tests and Interpreting Results
On this page
Overview
Paediatric diagnosis begins with stabilisation, history and examination — tests never precede resuscitation. An investigation is justified only when it answers a specific clinical question, changes management, or identifies a dangerous diagnosis that must not be missed. Children are not small adults: they have smaller circulating blood volumes (so indiscriminate sampling causes iatrogenic anaemia), rapidly changing age-dependent reference ranges, and different pre-test probabilities shaped by immunisation, malaria endemicity, HIV and malnutrition. Broad "fever panels" generate false positives, cost, parental anxiety and delayed treatment. The skilled clinician orders the smallest set of tests that safely distinguishes the urgent possibilities — and treats empirically when delay is more dangerous than uncertainty.
Learning objectives
- Formulate a problem representation and prioritised differential diagnosis before ordering any test.
- Select age-appropriate laboratory, microbiological and imaging investigations for common syndromes.
- Interpret results against age, gestation, nutrition, current treatment and local disease prevalence.
- Recognise when bedside testing, referral or empiric treatment beats extensive investigation.
- Communicate results, uncertainty, follow-up and safety-netting clearly to families.
Diagnostic reasoning sequence
- Stabilise first. ABCDE, glucose, oxygen, temperature and urgent treatment. No investigation justifies delaying resuscitation.
- Define the syndrome. Respiratory distress, fever without focus, diarrhoea/dehydration, seizure, jaundice, anaemia, malnutrition, rash, developmental concern — a named syndrome focuses the test list.
- Estimate the pre-test probability. Age, exposure history, immunisation status, malaria/TB/HIV prevalence, season and comorbidity decide which diagnoses are plausible enough to test for.
- Choose the focused test. Ask of every test: "what result would change what I do?" If nothing would, do not order it.
- Interpret in context. Compare with age-specific ranges, the child's trajectory, and the effects of current treatment (e.g. antibiotics before cultures, transfusion before the film).
- Act and reassess. Treat, observe, refer — or repeat testing only when the result would genuinely change management.
Core investigations
Scroll the table sideways on a smaller screen.
| Test | Questions it answers | Important cautions |
|---|---|---|
| Point-of-care glucose | Hypoglycaemia in the altered, shocked, malnourished or very young child | Treat severe low glucose immediately; confirm unexpected values but never delay treatment |
| Pulse oximetry | Hypoxaemia; need for oxygen and referral | Check probe fit, perfusion, motion and waveform; cold extremities mislead |
| Full blood count (FBC) | Anaemia, infection pattern (neutrophils, lymphocytes), platelets | Age-specific ranges (see next section); malaria, HIV and malnutrition distort every line |
| Malaria RDT / blood film | Malaria parasites where endemic; parasite count on film; morphology | A positive RDT reflects antigen, not viability — it can stay positive weeks after treatment; a negative test never excludes the other causes of fever |
| Urinalysis / culture | UTI, renal disease, dehydration, diabetes | Collection method dominates interpretation: bag and non-clean-catch samples are heavily contaminated; suprapubic or catheter specimens in infants |
| Blood culture | Sepsis/meningitis pathogen and antibiotic sensitivities | Take before antibiotics whenever it will not delay them; volume matters — under-filled bottles reduce yield |
| Lumbar puncture / CSF | Meningitis: cells, protein, glucose, Gram stain and culture | Contraindications first: signs of raised ICP, shock, severe respiratory compromise, bleeding disorder, local infection — or defer in a critically ill child and treat empirically |
| Electrolytes, urea, creatinine | Dehydration and its type (Na+), renal injury, drug safety | Interpret with fluid status, age and nutrition; hyponatraemia with CNS signs is dangerous |
| Liver function tests | Jaundice pattern (hepatitic vs cholestatic), drug toxicity | Direct vs indirect bilirubin separates haemolysis from liver disease; infants differ (physiological jaundice, immature conjugation) |
| CRP / procalcitonin | Bacterial vs viral likelihood; response to treatment | A support, not a verdict: early focal infection may have low CRP; interpret with the clinical picture |
| HIV test | Exposed infant (DNA/PCR) or child antibody testing with consent and counselling | Maternal antibodies invalidate antibody tests below ~18 months — use virological (PCR) testing in exposed infants |
Interpreting the full blood count
The FBC is the most-ordered and most-misread paediatric test. Read it line by line, always against age:
- Haemoglobin. Anaemia thresholds are age-specific (roughly < 11 g/dl in a 6-month-old, < 11.5 at 1–5 years, < 12 from 5 years). In malaria-endemic areas, a low Hb with a positive film is usually malarial anaemia; microcytosis suggests iron deficiency, macrocytosis folate/B12 or haemolysis.
- White cells and differential. Neonates normally have higher counts; lymphocytes dominate from ~6 months to 4 years (reverse of adults). High neutrophils with toxic granulation support bacterial infection; leukopenia with a left shift is ominous in sepsis; atypical lymphocytes suggest EBV; very high or very low counts demand consideration of leukaemia — but never diagnose leukaemia on a single count.
- Platelets. Thrombocytopenia in sepsis, malaria and dengue; reactive thrombocytosis after infection and iron deficiency. Count is less important than trend and bleeding.
- Never forget the film. The count tells you how much; the film tells you why — parasites, morphology, blast cells, fragments of microangiopathy.
Investigating common syndromes
Scroll the table sideways on a smaller screen.
| Syndrome | First-line approach | Escalate when |
|---|---|---|
| Fever without focus | Danger-sign screen; malaria RDT/film; urine testing in infants; blood culture if septic; HIV test in the exposed | Petechiae (meningococcaemia), persistent fever, hepatosplenomegaly, weight loss — think TB, typhoid, malignancy |
| Cough / fast or difficult breathing | Pulse oximetry is central; SpO2 and danger signs classify pneumonia severity; chest X-ray only for severe, atypical, persistent or complicated disease | Chest indrawing + SpO2 < 90%, suspected TB, foreign body, heart failure; no routine bloods in uncomplicated bronchiolitis |
| Diarrhoea / vomiting | Clinical dehydration classification first (none/some/severe); glucose in severe cases; stool only for blood, dysentery, outbreak or persistent diarrhoea | Shock, severe dehydration: electrolytes, urea, creatinine; suspected cholera for public health |
| Seizure | Glucose in every case; electrolytes including calcium and magnesium; the diagnosis is usually clinical | First prolonged or focal seizure, meningism, persisting altered state — treat, then LP when safe and contraindications excluded |
| Jaundice | Bilirubin (total and direct), FBC and film for haemolysis, liver enzymes; in the newborn use gestational-age/hour-specific bilirubin charts | Pale stools/dark urine (cholestasis), conjugated hyperbilirubinaemia, jaundice + encephalopathy |
| Malnutrition | MUAC and oedema first; FBC, electrolytes, glucose, HIV test, TB screen as indicated | Oedematous malnutrition (kwashiorkor) — watch electrolytes and hypoglycaemia closely |
Imaging and procedures
- Chest X-ray. Not a screening test: reserve for severe pneumonia, suspected TB, foreign body, heart failure, complicated or non-resolving disease. Portable AP films in sick children overestimate cardiomegaly and infiltrates.
- Ultrasound. First-line for abdominal pain and masses, pyloric stenosis (target/doughnut sign), intussusception (target sign), hip assessment in infants and procedural guidance; no radiation — the great advantage in children.
- CT and MRI. CT for trauma and suspected intracranial catastrophe; MRI for CNS detail, tumours and complex anatomy. Shield gonads, justify every scan — children are far more radiosensitive than adults, and a repeat scan is the commonest error.
- Procedures as investigations. Lumbar puncture (meningitis), bone marrow aspiration (marrow failure, leukaemia, TB), pleural or ascitic tap (sepsis, malignancy) — each with specific consent, analgesia and clear indications.
Reference ranges and test performance
Numbers without context mislead twice — once by being wrong for the age, once by being believed absolutely.
- Age-specific reference ranges. Haemoglobin, WBC differential, creatinine, bilirubin, CSF cell counts and protein all change with age and gestation; a "normal" adult range applied to a 2-month-old is a misreading. Neonates normally have higher WBC counts, higher CSF protein and more CSF cells than older children.
- Sensitivity. The proportion of diseased patients the test detects. A highly sensitive test rarely misses disease — use it to rule out (SnNout: Sensitive, Negative rules out).
- Specificity. The proportion of healthy patients correctly called negative. A specific test rarely labels healthy as diseased — use it to rule in (SpPin).
- Predictive values move with prevalence. PPV and NPV depend on pre-test probability, which is why the same malaria RDT means different things in a high-transmission district in April and a low-transmission area in December. This is the single most-examined concept in diagnostic testing.
Communicating results and safety-netting
- Explain before the result. Tell families what each test looks for and when results will come — an unexpected result without preparation breeds distrust and default from follow-up.
- State uncertainty honestly. "The test did not show malaria, but your child still has a fever, so if it worsens tonight, return immediately" is a safety-net, not a weakness.
- Give a written or verbal return plan. Where to go, when, with what warning signs — especially where phones and transport are barriers.
- Document and close the loop. Critical results (positive blood cultures, severe electrolyte derangement, HIV results) require read-back, action and a note of who was informed.
OSCE checklist
- Stabilise — ABCDE, glucose, oxygen before any test is discussed.
- Define the syndrome and state a ranked differential with pre-test probabilities.
- Justify each test aloud: the question it answers and the result that would change management.
- State cautions: age-specific ranges, treatment effects, volume limits, contraindications (LP, CT radiation).
- Interpret a given FBC/CSF/electrolyte result against the clinical picture.
- Communicate uncertainty, follow-up and the safety-net to the caregiver; document who was informed.
Worked case: two febrile children, two test strategies
Case A — 4-year-old, high-transmission district, rainy season. Fever for 2 days, looks reasonably well, drinks normally, no danger signs. Pre-test probability of malaria is high. One focused test: malaria RDT. Positive → artemether–lumefantrine per weight, counsel on completing the course, safety-net for warning signs. Negative → the fever has another cause: no treat-anyway reflex, review clinically, consider urine testing or watchful waiting with a clear return plan.
Case B — 28-day-old. The mother reports a "hot body", the baby feeds poorly and is lethargic. Pre-test probability of serious bacterial infection is high and every hour of delay costs organs. Do not let any test gate treatment: take the blood culture, malaria film and glucose, lumbar puncture if stable without contraindications, FBC — and give empiric antibiotics within the first hour. Here the tests refine therapy; they never delay it.
The contrast is the lesson. The same symptom, two completely different test strategies, decided by age and pre-test probability. In Case A the test rules a diagnosis in or out in an outpatient; in Case B the tests support treatment that must not wait. When you justify any test, name which case you are in.
Key points
- Tests answer questions — define the question before ordering, and stabilise before everything.
- Paediatric reference ranges are age-dependent; reading results against adult ranges is a systematic error.
- Sensitivity rules out, specificity rules in, and predictive values shift with prevalence — pre-test probability governs all interpretation.
- Syndrome-based selection (fever, respiratory, diarrhoea, seizure, jaundice, malnutrition) prevents both under- and over-investigation.
- Empiric treatment of a sick child beats a perfect test done too late; communicate uncertainty and a concrete return plan.
References and further reading
- Ministry of Health, Uganda. Uganda Clinical Guidelines — investigation standards for child health. (Library reference.)
- WHO. Integrated Management of Childhood Illness (IMCI) chart booklet — syndrome-based test selection. (Library reference.)
- Lissauer T, Fanaroff A, Miall L, Fanaroff J. Neonatology at a Glance — newborn reference ranges and bilirubin charts. (Library reference.)
- WHO. Laboratory diagnosis of malaria and RDT guidance. (Library reference.)
- Glasziou P, Irwig L, et al. Evidence-Based Medical Monitoring — sensitivity, specificity and predictive values in clinical use. (Library reference.)
For education only. Apply current Uganda Clinical Guidelines, local protocols and specialist advice in clinical practice.