In the United States, analysis of 13 years of national laboratory surveillance data from 1991 through 2004 established a characteristic spatiotemporal pattern of annual US rotavirus activity, with peak activity occurring first in the Southwest during the winter months and migrating across the continent to occur last in the Northeast approximately 2 to 3 months later.1,2 A recent study found that prevaccine spatiotemporal patterns of US rotavirus activity were correlated with geographical variations in birth rates but not with any environmental factors, the leading alternate hypothesis.3 Thus, Western US states with higher birth rates had earlier onset of rotavirus activity, suggesting that the regular annual epidemic dynamics were driven by the accumulation of fully susceptible individuals (ie, newborns) to the population. If this hypothesis was true, widespread vaccination would be expected to reduce the number of fully susceptible individuals, thereby delaying the timing of rotavirus activity and also potentially altering the seasonal pattern of rotavirus activity across the United States. To further investigate this hypothesis, we compared spatiotemporal trends in peak rotavirus activity during the most recent 2 surveillance years preceding rotavirus vaccine introduction in 2006 with those in the first 2 years following vaccine introduction.
MATERIALS AND METHODS
We analyzed rotavirus laboratory surveillance data from the National Respiratory and Enteric Viruses Surveillance System for July 2005 through June 2009.4 The National Respiratory and Enteric Viruses Surveillance System laboratories that reported testing results for a minimum of 30 weeks and conducted at least 100 rotavirus tests in each surveillance year (July through June) were included in the analysis. There were 55, 53, 47, and 46 laboratories meeting the eligibility criteria for each of the 4 surveillance years, respectively. To smooth the week-to-week variability in the number of positive rotavirus tests for individual laboratories, a 7-week central moving average of positive tests was calculated. If an eligible laboratory's central moving average of positive tests never exceeded 1 during the surveillance year, the laboratory was excluded from further analyses. This criterion resulted in the exclusion of 1, 2, 15, and 1 laboratory during the 4 surveillance years, respectively. The exclusion of 15 laboratories during the 2007 to 2008 surveillance year was consistent with the dramatically diminished and delayed rotavirus season previously described.5 For each year, we input the geographical location and the calendar month in which the maximum number of weekly rotavirus positive tests occurred for each laboratory into a Kriging linear spatial interpolation model (Surfer 8, Golden Software, Inc., Golden, CO) to determine the spatiotemporal patterns in rotavirus activity.6
The typical spatiotemporal pattern of US rotavirus activity was apparent during the last rotavirus season, before rotavirus vaccine introduction (2005–2006), and during 2006 to 2007, when coverage of rotavirus vaccine was very low and restricted to young infants (Fig. 1). In these years, peak rotavirus activity first occurred in the Southwest in January, then in the Midwest and West during February, and in the remainder of the country in March or April. In contrast, for both postvaccine introduction seasons during 2007 to 2009, the spatiotemporal pattern was considerably altered. During 2007 to 2008, most areas, including the Southwest, reached peak activity in April or May, approximately 2 to 3 months later than in prevaccine seasons. The spatiotemporal pattern was even more heterogeneous in 2008 to 2009, with rotavirus activity peaking as early as August in the extreme Northwest and localized peaks occurring during November through May with minimal national level pattern in activity. The early Northwest peak activity was due to the preceding delayed and prolonged rotavirus season in the local area.
Our findings illustrate that the 2 surveillance years following rotavirus vaccine introduction failed to demonstrate the typical Southwest to Northeast spread of US rotavirus activity observed consistently in 15 surveillance years preceding rotavirus introduction with available data. These remarkable alterations in US rotavirus activity have occurred rapidly following vaccine introduction and provide evidence that the spatiotemporal dynamics of rotavirus activity in the United States are likely driven by the rate of accumulation of fully susceptible individuals rather than solely by environmental factors.
1.Torok TJ, Kilgore PE, Clarke MJ, et al. Visualizing geographic and temporal trends in rotavirus activity in the United States, 1991 to 1996. National Respiratory and Enteric Virus Surveillance System Collaborating Laboratories. Pediatr Infect Dis J
2.Turcios RM, Curns AT, Holman RC, et al. Temporal and geographic trends of rotavirus activity in the United States, 1997–2004. Pediatr Infect Dis J
3.Pitzer VE, Viboud C, Simonsen L, et al. Demographic variability, vaccination, and the spatiotemporal dynamics of rotavirus epidemics. Science
4.Centers for Disease Control and Prevention. Reduction in rotavirus after vaccine introduction—United States, 2000–2009. MMWR Morb Mortal Wkly Rep.
5.Tate JE, Panozzo CA, Payne DC, et al. Decline and change in seasonality of US rotavirus activity after the introduction of rotavirus vaccine. Pediatrics
6.Waller LA, Gotway CA. Applied Spatial Statistics for Public Health Data.
Hoboken, NJ: John Wiley & Sons, Inc; 2004.