Page 53 - Read Online

Page 92 Swedberg et al. One Health Implement Res 2023;3:77-96 https://dx.doi.org/10.20517/ohir.2023.02

distribution of human cases likely arose from a combination of factors, including localized outbreaks,
inadequate dog vaccination coverage, suboptimal PEP-seeking behaviors, and potential variations in
surveillance and case detection capabilities. Mapping the locations of human cases alongside laboratory-
confirmed animal cases [Figure 4] clearly illustrated that nearly all animal rabies testing was conducted in
northern Oriental Mindoro in 2020 and 2021, whereas reported incidents of human deaths were limited to
the central and southern areas of the province. In 2022, sample collection increased throughout the
province; however, reported human cases remained in central and southern municipalities. The incidence
of high-risk bites showed a notable increase in 2022, likely due to the cancellation of MDV campaigns in
2020/2021 and the lifting of COVID-19 movement/travel restrictions, potentially leading to more exposure
events.

Our findings indicate that although human case detection is relatively robust, animal surveillance should be
enhanced to capture the incidence of rabies more effectively within the dog population. Over the 3-year
study, our decision tree model estimated a total of 1,678 rabid dogs (95% PrI 1,016-2,386) may have been
present in Oriental Mindoro. Yet only 28 animal cases were laboratory-confirmed during this time (case
detection of 1.7%). Notably, three municipalities accounted for more than half of positive dog cases (15 of
28), indicating stronger surveillance, though not necessarily a higher incidence of dog rabies. IBCM
surveillance protocols, which encouraged the investigation of suspected rabid animals and the collection of
samples in the case of dead or euthanized animals, led to a nearly fourfold increase in the detection of
laboratory-confirmed dog rabies cases from 2020 to 2022. However, external factors such as the COVID-19
pandemic and minimal to no dog vaccination in 2020 and 2021 make it difficult to discern whether the
higher case detection was exclusively due to surveillance being enhanced by IBCM or because of increased
rabies incidence within the dog population.

Strengths and limitations
We were typically able to classify biting animals as broadly “high-risk” or “low-risk” using initial patient risk
assessments from IBCM, but these did not always provide adequate information to differentiate between
WHO classifications “suspect” or “probable”. IBCM protocols specified risk assessments for every bite
patient presenting to ABTCs and investigations of any animal deemed high-risk. However, the COVID-19
pandemic and ensuing lockdowns contributed to challenges in the delivery of IBCM training and
subsequent implementation of protocols. Heavy workloads and temporary closure/reduced operating hours
of ABTCs limited the capacity of health workers to complete/submit risk assessments, while movement
restrictions prevented in-person animal investigations and affected sample collection. Challenges associated
with COVID-19 primarily affected IBCM implementation in 2020 and 2021, with 2022 mostly returning to
a relatively normal situation.

There was a potential bias towards the submission of high-risk bite data due to higher prioritization of
reporting, which may have resulted in overestimating rabies exposure incidence. However, attempts were
made to adjust for this by using nearly complete risk assessment data from one ABTC, located in the capital
city of Calapan, as well as the incomplete IBCM data submitted from all ABTCs, to extrapolate to the
province. Assuming there are differences in PEP-seeking behavior and endemicity of dog rabies between
urban and rural settings, both estimates come with limitations. However, these two methods of
extrapolation provide comparisons for our decision tree estimates and further evidence that only a small
percentage (< 3%) of bite patients seeking PEP were likely true rabies exposures.

Additional limitations include simplifying assumptions and uncertainties in our decision tree model
parameters. The parameters describing rabid dog biting (P bites|rabid_dog ) and the probability of infection