Background: Despite highly vaccinated populations in the United States, large-scale mumps outbreaks among adults continue to occur nationwide, often with painful patient outcomes and a heavy burden on public health resources. Most of the outbreaks have occurred in colleges/universities. Recent outbreaks have been identified in unique epidemiologically complex congregate settings/sub-populations. In these highly vaccinated populations, not recognizing mumps early as a potential cause of parotitis and not performing adequate laboratory testing of mumps (i.e., low sensitivity antibody IgM testing only vs. high sensitivity/specificity mumps RT-PCR) is limiting the understanding of the epidemiological factors involved in mumps outbreak propagation. After experiencing baseline mumps case counts on the average of ten per year in Los Angeles County (LAC), an unprecedented and large-scale multi-jurisdictional outbreak occurred in 2017 among males-who-have-sex-with-males (MSM) or individuals associating with MSM. A comparative analysis was performed of the mumps diagnostic omissions among outbreak and baseline/non-outbreak cases reported during the same time period. Objective: Explain the role that misdiagnosing mumps earlier in symptom presentation plays in outbreak propagation Methods: Demographic, clinical and laboratory testing data were analyzed among mumps cases reported to the LAC Department of Public Health during 2017. Cases classified as meeting the initial outbreak risk factors were compared to age range-matched non-outbreak cases on the following data elements: gender, occurrence of misdiagnosis and clinical presentation at the initial clinician evaluation, presence of co-morbidities like HIV, mumps immunization status prior to symptom development and mumps serology (IgM and IgG) and RT-PCR diagnostic testing. Results: From January-August 2017 (14 case generations), 50 confirmed LAC mumps cases were identified among MSM or MSM-associated individuals (age 19 to 60 years). Non-outbreak cases in this age group totaled 29 during the same time period. Eighty-four percent reported up-to-date vaccinations. Overall, 32% (n=16) of outbreak cases and 38% (n=11) of non-outbreak cases were initially misdiagnosed. Fifty percent of misdiagnosed outbreak cases would not have been diagnosed with IgM testing alone and without RT-PCR. Due in part to the lack of an accurate initial mumps diagnosis, the misdiagnosed outbreak cases exposed 11 persons in unique venues who subsequently became cases. Males comprised 88% of outbreak and 64% of non-outbreak misdiagnosed cases. Eighty percent (n=40) of outbreak and 93% (n=27) non-outbreak cases presented with parotitis, attributed to multiple etiologies with bacterial origin most commonly cited across both groups. Fifty-percent of female outbreak cases and 44% of female non-outbreak cases were misdiagnosed, with wide-ranging diagnoses in both groups. Conclusion: Mumps diagnostic omissions, both clinical and laboratory, impacted the length and characterization of this outbreak. When patients, regardless of immunization status, present with parotitis, inclusion of mumps in the differential diagnosis, alongside RT-PCR testing, assists in the identification of unique epidemiological factors that could be addressed earlier in curbing an outbreak.

Background: Since the 1990’s, increased rates of pertussis infection have been observed in Hispanic infants in the United States. During the 2010 epidemic in California, the rate among Hispanic infants <6 months of age was more than twice that of White infants. Waning efficacy of acellular pertussis vaccines, coupled with the rising proportion of Hispanics in the California population, make research on California pertussis epidemiology crucial. Few studies have explored the potential risk factors, which include >4 household contacts and urban residence. The majority of Hispanics in California are concentrated in the most urban areas, which could further explain over-representation of Hispanics among California’s infant pertussis cases if “urbanness” is a risk factor for pertussis. Objectives: To investigate whether living in a more densely populated area is a risk factor for infant pertussis in California, and whether any such association varies by ethnicity. Methods: Infant pertussis cases <6 months of age that were born in California and had cough onset between 2011 and 2014 were included in a case control study. Controls were matched on age and birth hospital. Participant addresses were geocoded and linked to 2010 U.S. Census tracts (CTs). CT population density greater than the study population median was used as a proxy for “urbanness” and was calculated using 2010 U.S. Census data. Odds ratios and 95% confidence intervals were calculated overall and stratified by ethnicity. Chi-squared tests and logistic regression were conducted with STATA 14.0. Results: A total of 399 pertussis cases and 1,168 controls were included in the analysis. Median CT population density was 6,765 persons per square mile. Urban infants had 1.43 times the odds (1.13-1.81) of pertussis compared to the non-urban group. Hispanic infants had 1.85 times the odds (1.43-2.40) of pertussis compared to non-Hispanic infants. When stratified by ethnicity, Hispanic infants in urban areas had a greater odds of pertussis compared to Hispanic infants in non-urban areas [OR=1.47 (1.10-1.97)], whereas “urbanness” was not significantly associated with pertussis among non-Hispanics [OR=1.01 (0.64-1.58)]. Using conditional logistic regression to control for ethnicity, there was a weak positive association between “urbanness” and pertussis [OR=1.30 (1.03-1.65)]. Conclusion: While living in more densely populated CTs was associated with a greater odds of infant pertussis in California, this association was modified by ethnicity. There was a significant, positive association between “urbanness” and pertussis infection for Hispanic infants but not for non-Hispanic infants, suggesting additional factors, such as household size, may be playing a role. Further investigation is needed into why Hispanic infants living in urban areas of California are at increased risk for pertussis.

Background: Seasonal influenza is a vaccine preventable disease that requires annual vaccination. In tropical countries, there are often two peaks of influenza activity—one in the middle and one at the end of the year, of variable magnitude. In these countries, it can be difficult to determine the optimal time for annual vaccine application. We examined patterns of influenza circulation in five tropical countries in the Americas using two different mathematical approaches, in order to establish the timing of the primary and secondary influenza seasons and hence the optimal period for vaccination. Objectives: The objective of this analysis was to determine the timing of the peaks in influenza activity and hence determine the ideal timing of the application of the influenza vaccine. Methods: Virologic data on the number of surveillance samples tested and the number of samples positive for influenza were obtained from the open-access World Health Organization (WHO) FluNet database for five countries—Colombia, Costa Rica, Cuba, Ecuador, and El Salvador for the period from 2010-2016. First, the percent of surveillance samples positive for influenza among all samples tested was calculated by epidemiologic week (EW) of sample collection. Second, using the time series of data, an averaging method and a negative binomial regression model were each used to establish expected percent positivity for influenza among the samples tested, by EW during the time series and the results from each approach were compared. Lastly, we examined the timing of the current vaccination campaign relative to the timing of the largest peak in activity to look for mismatched vaccine application. The negative binomial regression analysis was conducted in R Studio and the averaging analysis was conducted using Microsoft Excel. Results: There were a total of n=178,798 samples tested for influenza in these five countries during the period of analysis. All countries displayed two peaks in the influenza percent positivity each year. The timing of the primary and secondary seasons, using the negative binomial model and the averaging method, were comparable, while the magnitude of the peaks differed. Among the five countries, three are currently vaccinating against influenza before their largest peak in activity. Conclusion: We used two different mathematical approaches to establish the seasonality of influenza in tropical countries and found the timing of the seasons to be comparable. Tropical countries with two influenza seasons should consider using historic data, such as percent positivity for influenza among all surveillance samples tested, to establish their country’s seasonality and hence the optimal timing for the application of the vaccine.