The impact of the COVID-19 pandemic on rabies reemergence in Latin America: The case of Arequipa, Peru

Abstract

In Latin America, there has been tremendous progress towards eliminating canine rabies. Major components of rabies elimination programs leading to these successes have been constant and regular surveillance for rabid dogs and uninterrupted yearly mass dog vaccination campaigns. Unfortunately, vital measures to control COVID-19 have had the negative trade-off of jeopardizing these rabies elimination and prevention activities. We aimed to assess the effect of interrupting canine rabies surveillance and mass dog vaccination campaigns on rabies trends. We built a deterministic compartment model of dog rabies dynamics to create a conceptual framework for how different disruptions may affect rabies virus transmission. We parameterized the model for conditions found in Arequipa, Peru, a city with active rabies virus transmission. We examined our results over a range of plausible values for R0 (1.36-2.0). Also, we prospectively evaluated surveillance data during the pandemic to detect temporal changes. Our model suggests that a decrease in canine vaccination coverage as well as decreased surveillance could lead to a sharp rise in canine rabies within months. These results were consistent over all plausible values of R0. Surveillance data from late 2020 and early 2021 confirms that in Arequipa, Peru, rabies cases are on an increasing trajectory. The rising rabies trends in Arequipa, if indicative to the region as whole, suggest that the achievements made in Latin America towards the elimination of dog-mediated human rabies may be in jeopardy.

Fig 1. Rabies compartmental model results.

Simulations depict daily point prevalence of each disease state (S,E,I,V). Panel A shows the dynamics of all disease states in the best fit rabies model for Arequipa, Peru. The blue line shows the vaccinated dog population numbers over time and the yellow line shows the susceptible population dynamics. Because the proportion of rabies exposed (pink line) and infected dogs (red line) is so small, these dynamics are not apparent in Panel A. Panel B highlights these exposed and infectious dynamics with an adjusted scale. Panel C shows infected population dynamics for a range of R₀ from 1.36–1.5. In other words, it represents the red line of infected population dynamics shown in B but for a range of R₀. The trends extend for the full range of possible values of R₀ [1.36, 2.0] which can be seen in the supplementary information (Figure D in S2 Appendix).

Fig 2. Different simulations of disruption scenarios for 1 year.

Simulations depict daily point prevalence of each disease state (S,E,I,V). Simulations were run for 1 year after the beginning of COVID-19 control measures in Arequipa, Peru (March 16, 2020- March 16, 2021). Panels A-C depict different vaccination scenarios with normal levels of surveillance and control measures leading to an average survival time (ST) of rabid dogs to be 2.5 days. Panels D-F show the same vaccination scenarios with decreased surveillance leading to an increased survival time of rabid dogs to 3.7 days. The vaccination scenarios depicted correspond to yearly vaccination campaigns reaching the optimal goal of 80% coverage (Panels A, D), a sub-optimal level of 58% coverage (Panels B, E), and a complete cancellation of the vaccination campaign where coverage is 0% (Panels C, F). Both the surface plots with a range of values of R₀ (i) and a transect where R₀ = 1.44 (ii) are displayed).

Fig 3. Canine rabies diagnostic samples submitted from 2015–2021.

Averages for monthly canine rabies case counts (monthly cumulative incidence) confirmed by laboratory diagnosis (Panel A) and for total canine diagnostic sample submissions to the Peru Ministry of Health for Arequipa Department (Panel B) were calculated by aggregating monthly data from March 2015 (when rabies was first reported in Arequipa city) to February 2020 (the last month that Peru was operating under non-COVID conditions) and compared to pandemic time surveillance data (March 2020 to March 2021). 95% confidence intervals were computed via bootstrap methods by resampling 2000 times. A full timeline is available in the supplementary information (Figure C in S2 Appendix).

Fig 4. Disruption scenario simulations for 5 years.

Rabies dynamics in Arequipa were simulated for five years under four different scenarios. The “No disruption” scenario depicts a counterfactual no-covid scenario where vaccination campaigns were conducted at pre-pandemic normal (58% coverage) and average surveillance remained normal (average survival time of rabid dogs = 2.5 days) throughout the simulated time period (March 2020-March 2025). The disruption scenarios simulate 1–3 years of COVID-19 disruptions with vaccination coverage reduced to 12.3% and average survival time of rabid dogs increased to (3.7 days) before returning to pre-pandemic normal.

Fig 5. Map of Arequipa Department.

Arequipa Department consists of 8 provinces. Prior to February 2021, rabies cases were contained in Arequipa province (the province containing Arequipa city). In February 2021, cases spread to the neighboring province of Caylloma, primarily in the city of El Pedregal. Seven cases have since been detected in El Pedregal. Shapefiles used to create maps are from Peru’s National Geo-referenced Data Platform Geo Peru [32].