BACKGROUND: Real-time RT-PCR can detect low levels of norovirus in stool specimens from symptomatic cases as well as asymptomatic or previously resolved infections, so the clinical interpretation is sometimes unclear. The cycle threshold (Ct) value of real-time RT-PCR, which is inversely proportional to viral load, may be useful to improve clinical diagnosis. We therefore evaluate how demographic, clinical, and outbreak characteristics are associated with norovirus Ct values in diagnostic specimens.
METHODS: Laboratory and epidemiological data were extracted from the following databases which included outbreaks, sporadic cases, and healthy controls: (1) the U.S. national norovirus outbreak surveillance network (CaliciNet, n=5997), (2) the active pediatric acute gastroenteritis surveillance system in seven U.S. medical institutions (New Vaccine Surveillance Network, n=588), and (3) population-based studies in Latin America (n=197). To assess relationships between norovirus Ct values and each factor (sex, age, case-control status, norovirus genogroup, month of symptom onset, timing of specimen collection, transmission mode, outbreak setting, and rotavirus testing result) while controlling for others, we fit mixed-effect Poisson regression models including the study as a random effect and other factors as fixed effects. We also performed receiver-operating characteristic analysis with the aim of defining a cut-off in the Ct values to differentiate symptomatic cases from asymptomatic cases.
RESULTS: According to regression models, norovirus Ct values were lower among symptomatic cases (mean ± standard error: 25.3±1.2) compared to asymptomatic cases (28.6±1.4, p<.0001) after controlling for other factors. Specimens from young children and the elderly had lower Ct values (25.9−27.0) than those in the middle age group (28.0−28.7, p<.0001). Norovirus genogroup (G) II specimens had lower Ct values (26.2±1.1) than those of GI (27.4±1.1, p<.0001). Ct values were higher when specimens were collected >7 days after illness onset (27.1−29.4), compared to those collected within 7 days (24.2−24.7, p<.0001). Outbreaks transmitted by food or unknown/other modes had lower Ct values (both 23.6±0.6) than outbreaks with person-to-person transmission (24.3±0.6, p=0.002). Among all settings, cruise ship outbreaks had the lowest Ct values (21.8±0.9). Optimal Ct cut-off was 27 for GI and 26 for GII, but sensitivity and specificity were low.
CONCLUSIONS: Ct values were lower among symptomatic cases compared with healthy controls, and were related to age, norovirus genogroup, timing of specimen collection, transmission mode, and outbreak setting. Ct values may not serve as an effective cutoff to discriminate between cases and controls, but these characteristics could be taken into account for evaluating disease burden of norovirus.