Seroprevalence and Risk Factors for Rickettsia and Leptospira Infection in Four Ecologically Distinct Regions of Peru

Rickettsia and Leptospira spp. are under-recognized causes of acute febrile disease worldwide. Rickettsia species are often placed into the spotted fever group rickettsiae (SFGR) and typhus group rickettsiae (TGR). We explored the antibody prevalence among humans for these two groups of rickettsiae in four regions of Peru (Lima, Cusco, Puerto Maldonado, and Tumbes) and for Leptospira spp. in Puerto Maldonado and Tumbes. We also assessed risk factors for seropositivity and collected serum samples and ectoparasites from peri-domestic animals from households in sites with high human seroprevalence. In total, we tested 2,165 human sera for antibodies (IgG) against SFGR and TGR by ELISA and for antibodies against Leptospira by a microscopic agglutination test. Overall, human antibody prevalence across the four sites was 10.6% for SFGR (ranging from 6.2% to 14.0%, highest in Tumbes) and 3.3% for TGR (ranging from 2.6% to 6.4%, highest in Puerto Maldonado). Factors associated with seroreactivity against SFGR were male gender, older age, contact with backyard birds, and working in agriculture or with livestock. However, exposure to any kind of animal within the household decreased the odds ratio by half. Age was the only variable associated with higher TGR seroprevalence. The prevalence of Leptospira was 11.3% in Puerto Maldonado and 5.8% in Tumbes, with a borderline association with keeping animals in the household. We tested animal sera for Leptospira and conducted polymerase chain reaction (PCR) to detect Rickettsia species among ectoparasites collected from domestic animals in 63 households of seropositive participants and controls. We did not find any association between animal infection and human serostatus.

Burden of Influenza in 4 Ecologically Distinct Regions of Peru: Household Active Surveillance of a Community Cohort, 2009-2015

BACKGROUND

There are limited data on the burden of disease posed by influenza in low- and middle-income countries. Furthermore, most estimates of influenza disease burden worldwide rely on passive sentinel surveillance at health clinics and hospitals that lack accurate population denominators.

METHODS

We documented influenza incidence, seasonality, health-system utilization with influenza illness, and vaccination coverage through active community-based surveillance in 4 ecologically distinct regions of Peru over 6 years. Approximately 7200 people in 1500 randomly selected households were visited 3 times per week. Naso- and oropharyngeal swabs were collected from persons with influenza-like illness and tested for influenza virus by real-time reverse-transcription polymerase chain reaction.

RESULTS

We followed participants for 35353 person-years (PY). The overall incidence of influenza was 100 per 1000 PY (95% confidence interval [CI], 97-104) and was highest in children aged 2-4 years (256/1000 PY [95% CI, 236-277]). Seasonal incidence trends were similar across sites, with 61% of annual influenza cases occurring during the austral winter (May-September). Of all participants, 44 per 1000 PY (95% CI, 42-46) sought medical care, 0.7 per 1000 PY (95% CI, 0.4-1.0) were hospitalized, and 1 person died (2.8/100000 PY). Influenza vaccine coverage was 27% among children aged 6-23 months and 26% among persons aged ≥65 years.

CONCLUSIONS

Our results indicate that 1 in 10 persons develops influenza each year in Peru, with the highest incidence in young children. Active community-based surveillance allows for a better understanding of the true burden and seasonality of disease that is essential to plan the optimal target groups, timing, and cost of national influenza vaccination programs.