Fever in Pakistan: dengue, typhoid or malaria — and what to test first

A patient comes in with four days of fever, no obvious focus, and looks unwell but not dying. In Pakistan that is one of the most common presentations there is — and one of the few where the honest first answer is "I don't know yet, and I shouldn't pretend otherwise."

Dengue, typhoid and malaria overlap almost completely in the first week. The classical teaching — rose spots, retro-orbital pain, tertian fever — describes textbook cases, and textbook cases are the minority. What actually protects the patient is a disciplined order of operations, not a confident guess.

Reorder the question

The instinct is to ask "which of the three is this?" That's the wrong question first, because two things matter more:

1. Is this patient sick enough to need resuscitation now? If yes, the label matters less than the response — and several of the danger features are shared anyway.
2. Have I ruled out the one that kills fastest? Falciparum malaria can kill within days.

Only after those does differentiating help. That ordering is the whole point of this article.

Red flags first

Severe malaria

WHO defines severe falciparum malaria as one or more of the following, in the absence of an identified alternative cause and in the presence of asexual parasitaemia:

• Impaired consciousness — Glasgow coma score < 11 in adults, or Blantyre coma score < 3 in children
• Prostration — unable to sit, stand or walk without assistance
• Multiple convulsions — more than two in 24 hours
• Acidosis — base deficit > 8 mEq/L, bicarbonate < 15 mmol/L, or lactate ≥ 5 mmol/L; clinically, rapid deep laboured breathing
• Hypoglycaemia — glucose < 2.2 mmol/L (< 40 mg/dL)
• Severe anaemia — Hb ≤ 5 g/dL or Hct ≤ 15% in children under 12 (< 7 g/dL, < 20% in adults), with parasites > 10 000/µL
• Renal impairment — creatinine > 265 µmol/L (3 mg/dL) or urea > 20 mmol/L
• Jaundice — bilirubin > 50 µmol/L (3 mg/dL) with parasites > 100 000/µL
• Pulmonary oedema — radiological, or SpO₂ < 92% on air with respiratory rate > 30/min
• Significant bleeding, shock, or hyperparasitaemia > 10%

Severe vivax and knowlesi malaria: defined as for falciparum malaria but with no parasite density thresholds.

That single line deserves its own paragraph, because P. vivax predominates in Pakistan. A vivax patient with jaundice, or with severe anaemia, is severe malaria — they do not need to meet the falciparum parasite counts. Vivax being treated as the "benign" malaria is one of the more dangerous habits in this setting.

Severe dengue

Severe plasma leakage causing shock or fluid accumulation with respiratory distress; severe bleeding; or severe organ involvement — AST or ALT ≥ 1000, impaired consciousness, or cardiac and other organ involvement.

Complicated typhoid

Jaundice, drowsiness, severe abdominal pain, intestinal haemorrhage, severe sepsis or septic shock. MMIDSP directs that these patients be managed at a tertiary centre.

Notice how much these three lists share: jaundice, impaired consciousness, bleeding, shock. A patient with any of them needs resuscitation and referral regardless of which organism is responsible — which is exactly why the label can wait.

What actually shifts the odds

Once red flags are excluded, these features narrow the differential. They do not settle it.

What to test, in what order

1. Malaria film or RDT — always, first

Not because it's most likely, but because it's fast, cheap, and the one that kills quickest if missed. Test every patient with undifferentiated fever in an endemic setting, however dengue-like the picture looks — and repeat it if the first is negative and suspicion persists. A single negative film does not exclude malaria.

2. Two sets of blood culture — before antibiotics

MMIDSP: two sets are optimal before starting antibiotic therapy, and also in patients already on antibiotics who aren't responding. Cultures can flag positive as early as 4 hours, with identification and susceptibility within about 48.

That susceptibility result is not a formality in Pakistan — it determines the entire treatment path, because it separates drug-sensitive from MDR from XDR from ESBL typhoid, and those get four different regimens. Starting antibiotics before culturing throws that away.

3. Full blood count with serial haematocrit

The WHO dengue laboratory warning sign is a rising haematocrit with a concurrent rapid fall in platelets — the combination and the trend, not one number in isolation. Leukopenia supports dengue; bicytopenia is a typhoid complication marker.

4. Dengue NS1 and/or serology, read against the day of illness

NS1 is highest early; antibodies appear later. A test interpreted without reference to illness day is easy to over- or under-call.

The test not to order

Positive serological tests (such as Widal and TyphiDOT) are not recommended for diagnosis of enteric fever and are not included in case definitions of typhoid. Serological tests should never be ordered or relied upon to diagnose or rule out enteric fever.

That is Pakistan's own national society guidance, not an imported opinion. Widal remains in daily use here and it generates both false reassurance and false diagnoses — a positive result in an endemic population means very little, and a negative one excludes nothing.

Three traps worth naming

The dengue patient who is "improving". Deterioration in dengue classically happens as the fever settles, not at its peak. A patient looking better on day four may be entering the critical phase. Warn the family explicitly: the fever coming down is not the same as getting better.

The typhoid patient still febrile on day three. Defervescence in typhoid is prolonged and may take five to seven days. MMIDSP warns specifically against rushing to change antibiotics, and notes appetite and general condition often improve before the fever does. A patient eating and sitting up while still spiking is treatment working.

Assuming one diagnosis excludes the others. Co-infection happens. A positive malaria film in a patient with a falling platelet count does not mean dengue isn't also present.

Bottom line

Ask "is this patient sick?" before "which one is it?". Test for malaria in everyone, first, and repeat if negative. Send two blood cultures before any antibiotic. Read platelets alongside haematocrit and against yesterday's value. Don't order Widal. Remember that severe vivax needs no parasite threshold. And watch the dengue patient hardest at the moment they appear to be improving.

Our fever tool implements this order of operations — red flags first, then the features that shift the odds, then the testing sequence — and routes into the dengue and typhoid wizards once the picture clarifies. It deliberately never lets malaria be excluded on clinical features.

ReviseFCPS1 is built by doctors in Pakistan, for doctors in Pakistan. All our clinical tools are free. If you are preparing for FCPS Part 1, try it free for 7 days — 10,176 doctor-verified MCQs and 8,236 real exam recalls.