What serology tests are recommended for suspected whooping cough (pertussis)?

Serology Testing for Suspected Whooping Cough

For suspected pertussis, order paired acute and convalescent sera to measure IgG antibodies to pertussis toxin (PT), with the first sample taken within 2 weeks of cough onset and the second 3-4 weeks later; however, serology should not be your primary diagnostic test—nasopharyngeal culture or PCR are far superior for confirming the diagnosis. 1, 2

Primary Diagnostic Approach (Not Serology)

Before discussing serology, understand that serology is not the recommended first-line test for pertussis diagnosis:

• Nasopharyngeal culture remains the gold standard (100% specific) and should be your initial diagnostic test 1, 2
• PCR testing is preferred when clinical criteria are met (cough >2 weeks with paroxysms, whooping, or post-tussive vomiting), offering 80-100% sensitivity compared to culture's 25-50% 1, 2
• The CDC does not endorse single-sample serologic testing for routine diagnostic use because these assays lack standardization and cannot differentiate between recent disease, remote disease, or vaccination 3

When to Use Serology

Serology has a limited but specific role:

Paired Sera Testing (The Only Valid Serologic Approach)

• Order paired acute and convalescent sera showing a fourfold increase in IgG or IgA antibodies to PT or filamentous hemagglutinin (FHA) 1
• Timing is critical: First sample within 2 weeks of cough onset, second sample 3-4 weeks later 1
• This approach has 99% specificity but only 63% sensitivity for documenting pertussis 1

Specific Serologic Tests to Order

• IgG antibody to pertussis toxin (PT) via enzyme-linked immunosorbent assay (ELISA) is the most generally accepted serologic criterion 1
• IgA antibodies to PT or FHA can also be measured as part of the paired sera approach 1
• Additional antigens available include pertactin and fimbriae, though PT remains the primary target 1

Critical Limitations of Serology

Why Serology Often Fails in Clinical Practice

• Patients delay seeking care, making it impossible to obtain properly timed paired samples 1
• Non-rising titers may represent past infection or previous immunization, not acute disease 1
• No serologic method has been validated and approved for diagnostic use in the United States 1
• Single serum specimens showing high titers are only "highly suggestive" when there's a compatible clinical picture, but this approach lacks validation 1

Sensitivity Issues

• Acute serum samples (1-3 weeks after cough onset) show extremely low sensitivity of only 2-19% 4
• Even convalescent samples (5-10 weeks after symptom onset) achieve only 61-74% sensitivity for IgG anti-PT 4
• Combining IgG anti-PT with IgA anti-FHA improves sensitivity to 81-89% in convalescent samples, but this is still inferior to PCR 4

Practical Clinical Algorithm

When you suspect pertussis (cough >2 weeks with paroxysms, whooping, or post-tussive vomiting):

1. First-line: Order nasopharyngeal aspirate or Dacron swab for culture 1, 2
2. Simultaneously: Order PCR if clinical case definition is met (significantly more sensitive than culture) 2
3. Do not delay treatment while awaiting results—start macrolide antibiotics immediately 2, 5
4. Consider paired sera only if:
   • Patient presents early enough (within 2 weeks of cough onset) to obtain properly timed samples 1
   • You need retrospective confirmation for epidemiologic purposes 1
   • Culture and PCR are negative but clinical suspicion remains high 1

Common Pitfalls to Avoid

• Do not order single-sample serology expecting a definitive diagnosis—it cannot distinguish recent infection from past infection or vaccination 3
• Do not wait for serologic results to initiate treatment, as early treatment (within first 2 weeks) is critical for reducing transmission and coughing paroxysms 2
• Do not rely on serology after 3 weeks of symptoms, as culture sensitivity drops to 1-3% and the diagnostic window for optimal testing has passed 2
• Remember that 80% of culture-confirmed cases show significant antibody changes, but 20% may not mount a detectable serologic response 6