Correlation of Real Time PCR Cycle Threshold Cut-Off with Bordetella pertussis Clinical Severity

Bordetella pertussis testing performed using real-time polymerase chain reaction (RT-PCR) is interpreted based on a cycle threshold (Ct) value. At Public Health Ontario Laboratories (PHOL), a Ct value <36 is reported as positive, and Ct values >= 36 and <40 are reported as indeterminate. PHOL reported indeterminate results to physicians and public health units until May 2012, after which these results were only reported to physicians. We investigated the association between Ct value and disease symptom and severity to examine the significance of indeterminate results clinically, epidemiologically and for public health reporting. B. pertussis positive and indeterminate RT-PCR results were linked to pertussis cases reported in the provincial Integrated Public Health Information System (iPHIS), using deterministic linkage. Patients with positive RT-PCR results had a lower median age of 10.8 years compared to 12.0 years for patients with indeterminate results (p = 0.24). Hospitalized patients had significantly lower Ct values than non-hospitalized patients (median Ct values of 20.7 vs. 31.6, p<0.001). The proportion of patients reporting the most indicative symptoms of pertussis did not differ between patients with positive vs. indeterminate RT-PCR results. Taking the most indicative symptoms of pertussis as the gold-standard, the positive predictive value of the RT-PCR test was 68.1%. RT-PCR test results should be interpreted in the context of the clinical symptoms, age, vaccination status, prevalence, and other factors. Further information on interpretation of indeterminate RT-PCR results may be needed, and the utility of reporting to public health practitioners should be re-evaluated.