Transmission dynamics and vaccination strategies for Crimean-Congo haemorrhagic fever virus in Afghanistan: A modelling study.

BACKGROUND: Crimean-Congo haemorrhagic fever virus (CCHFV) is a highly pathogenic virus for which a safe and effective vaccine is not yet available, despite being considered a priority emerging pathogen. Understanding transmission patterns and the use of potential effective vaccines are central elements of the future plan against this infection.
METHODS: We developed a series of models of transmission amongst livestock, and spillover infection into humans. We use real-world human and animal data from a CCHFV endemic area in Afghanistan (Herat) to calibrate our models. We assess the value of environmental drivers as proxy indicators of vector activity, and select the best model using deviance information criteria. Finally we assess the impact of vaccination by simulating campaigns targeted to humans or livestock, and to high-risk subpopulations (i.e, farmers).
FINDINGS: Saturation deficit is the indicator that better explains tick activity trends in Herat. Recent increments in reported CCHFV cases in this area are more likely explained by increased surveillance capacity instead of changes in the background transmission dynamics. Modelling suggests that clinical cases only represent 31% (95% CrI 28%-33%) of total infections in this area. Vaccination campaigns targeting humans would result in a much larger impact than livestock vaccination (266 vs 31 clinical cases averted respectively) and a more efficient option when assessed in courses per case averted (35 vs 431 respectively). Targeted vaccination of farmers is impactful and more efficient, resulting in 19 courses per case averted (95% CrI 7-62) compared to targeting the general population (35 courses 95% CrI 16-107).
CONCLUSIONS: CCHFV is endemic in Herat, and transmission cycles are well predicted by environmental drivers like saturation deficit. Vaccinating humans is likely to be more efficient and impactful than animals, and importantly targeted interventions to high risk groups like farmers can offer a more efficient approach to vaccine roll-out.

Introduction

Crimean-Congo haemorrhagic fever virus (CCHFV) is an emerging tick-borne zoonotic pathogen which can lead to cases of fatal haemorrhagic fever in humans. In recent years, outbreaks of CCHF in humans have increased in frequency, and the virus is now endemic in several countries in the Middle East, Africa, Asia, and Southeast Europe. The wide geographical distribution of tick species which are able to harbour the virus provides added concern that the disease may spread further afield. This, and CCHF’s epidemic-proneness has led the World Health Organisation to include CCHFV in the group of priority pathogens for research and development into improved vaccines, therapeutics and diagnostics [1].

The transmission dynamics of CCHFV is complex due to the interplay between environmental factors affecting tick activity and their life cycles, the asymptomatic transmission within multiple vertebrate species (wild and livestock), and behavioural factors behind the risk of spillover into humans. Hyalomma spp. are the main vectors of CCHFV, and Hyalomma marginatum complex the most frequently associated species. Tick activity has been associated with environmental and meteorological variations, which might drive seasonal transmission patterns [2]. Hyalomma spp, specifically, thrives during the hot summer months in dry weathers, however its adaptability to colder temperatures has been also reported, explaining in part the expanding geographical area of influence of CCHFV [2-4]. Emergence of CCHFV has also been linked to importation of livestock species [5] and changes in agricultural activities, which affect the habitats of intermediate hosts of CCHFV [2,6].

Despite the absence of a safe and effective licensed vaccine against CCHFV, the development of a stable animal model [7] has meant that several vaccine candidates are now being studied. Inactivated virus [8,9], DNA [10], mRNA [11], and plant-expressed glycoprotein formulations [12] amongst others, are part of the current development pipeline. Inactivated vaccines have been routinely used in humans before in Bulgaria [8], with reported reductions in incidence, but the lack of data on efficacy and safety of this formulation has prevented its wider use.

CCHFV is considered a priority emerging pathogen, but important gaps in our understanding of transmission dynamics into humans and a formal assessment of the potential impact of vaccines are still necessary to advance a global research agenda and for developing a roadmap for CCHF.

Human CCHFV cases have been reported in Afghanistan at least since 1998, first in Takhar province, and later in Herat province where most of the cases have emerged since 2002 [13]. However, in recent years the distribution of cases has extended to most provinces in the country. Neighbouring countries such as Pakistan, Iran, Turkmenistan, and Tajikistan, also report annual cases of CCHFV. These countries are located in the ecological range of activity of Hyalomma spp. Transboundary livestock movement is thought to aid transmission [14].

Here we present a first approach to modelling CCHFV transmission amongst livestock and from livestock into humans, in Herat, an endemic area of Afghanistan.

Herat reported CCHFV outbreaks in 2008 and 2017 with an estimated case fatality ratio (CFR) ranging from 22% to 33% in humans [15,16]. Seroprevalence studies carried out in the area have reported high IgG seroprevalence in livestock (~75%) and higher seroprevalence among humans involved in livestock activities (farming, animal husbandry, etc) [17]. High seroprevalence in livestock suggests endemic transmission in animal hosts, although further serological evidence is not available to confirm this. Furthermore, the drivers behind trends in human spillover are not fully understood and the possible seasonality driven by environmental factors has not been assessed yet.

In this work, we aim to shed light on the main factors driving CCHFV transmission in western Afghanistan and ascertain whether the disease is endemic or epidemic, as a case study for CCHFV in general. We expand this case to explore the impact of selected vaccination strategies on disease incidence and mortality reduction in humans.

Methods

Ethics statement

This study has obtained approval from the ethics committee at the London School of Hygiene and Tropical Medicine (Reference number: 26612). All the data used has been aggregated and anonymised.

Mathematical model

We model CCHFV transmission in livestock, and from livestock to humans in two steps ().

Model schematic of CCHFV transmission.

We modelled CCHFV viral transmission between livestock, and from livestock to humans. In panel A, livestock were stratified in five yearly age-groups. Animals are born into the model at a rate proportional to mortality to maintain equilibrium. A fraction of livestock will acquire immunity through colostrum exposure in the first days after birth, here denoted as compartment Ri, the remaining fraction will enter the model through S1. The fraction moving to Ri is proportional to CCHFV prevalence at each time t. We assume a colostrum acquired immunity loss after six months. We model transmission between livestock with a risk function that renders the expected tick activity as a function of environmental drivers, and incorporates CCHFV infection prevalence in livestock and a scaling factor for the climatic indicator.. Infectious animals (I) recover after seven days, on average passing into the recovery compartment (R). We assume waning immunity with an average rate of 5 years-1, hence the transition R -> S. (B), we formulate the human spillover structure as an SEIRS model, governed by a series of stochastic transition events. Transmission follows a force of infection λ, that is defined by the infectious livestock prevalence at time t and a relative risk of transmission that conveys the differential risk by human occupation (i.e., farmers, and other). This connecting link is represented by the red dashed arrow connecting the two models. This implies a sequence of events in the runtime in which a realisation of the animal model is run over the time period, producing a vector of prevalence as output. This vector is subsequently passed as an input to the human spillover model. From the structure in panel B, is also evident that we allow a loss of acquired immunity R->S.

In step 1, we define a deterministic susceptible-infected-recovered-susceptible (SIRS) model structure for livestock, stratified in five yearly age groups (see ). Transmission between animals occurs as a function of prevalence of infectious livestock and a driving environmental factor (e.g., saturation deficit, soil temperature) which acts as a surrogate indicator of tick activity.. Hence in livestock, we define the force of infection as following,

Where the term represents the prevalence of infectious livestock at any given point in time, while β is the transmission coefficient representing the combination of transmission likelihoods between tick and livestock, and tick activity. We can also write,

Where R(t) is the reproduction number in livestock at each point in time t, and is defined as a function of the environmental driver used; D is the duration of the infectious period in livestock. Since environmental drivers are here reflecting a measure of tick activity, we incorporate the conditions that best reflect tick activity in relation to each driver. For a temperature dependent reproduction number, for example, we describe a system where adult Hyalomma spp. activity occurs above 12°C [2,18] and increases as temperature increases. Once temperature reaches above 30°C, ticks prefer to bury into soil [19], thus we write a function for declining transmission.

In step 2, we use a stochastic susceptible-exposed-infected-recovered-susceptible (SEIRS) (see ) model for transmission of CCHFV from livestock into humans. We define the process of transmission with a random binomial process where probability of event depends on infection prevalence in livestock at each time t and a risk multiplication factor to capture excess risk among farmers (assumed to be the high risk group). Human to human transmission is not modelled, assuming that this component is not relevant in sustaining CCHFV outbreaks. See and for a full model description and model equations.

The model is fitted to data by calibrating relevant model parameters within a Bayesian framework (see calibration procedures in ). In , we present the list of model parameters and the calibrated values. We compare our model output against target data on age stratified CCHFV seroprevalence in livestock, risk stratified seroprevalence in humans (i.e., farmers and other occupations), and time series of reported CCHFV cases in humans (Table A in ). Calibration diagnosis can be found in Figs A-C in

Exploring epidemiological and environmental drivers

In the absence of tick activity data, we use the environmental factors that influence tick dynamics and its trends over the year. These factors are incorporated as drivers of transmission of CCHFV between livestock. To explore the different environmental drivers and potential epidemiological conditions that better explain the observed trends in Herat’s data, we systematically compare models in two steps.

In step 1, we calibrate the models four times, each time using a different environmental driver, namely soil temperature, saturation deficit, relative humidity and normalized difference vegetation index (NDVI).

We retrieved the relevant data for the specific geographical location and time period from available sources. Each driver, with its source and relevance in tick activity can be found in (For further details on construction of these indicators see ).

In step 2, we further explore assumptions about the epidemiological factors behind the trends in reported human CCHF cases over the years. For this, we use the best model selected in step 1, and test three potential scenarios that have been hypothesised elsewhere [13,15], namely: A) Increments in CCHF reported cases reflect increment in reporting capacity (baseline assumption); B) Increased influx of livestock from other endemic regions (with a fixed reporting capacity); and, C) Increased influx of livestock from other endemic regions, and increased reporting capacity combined.

For livestock, we assume wanning immunity over an average period of 5 years after infection. This assumption allows us a better model calibration when compared against the lifelong immunity assumption (see and ).

In both steps 1 and 2, the most appropriate model is selected using a Deviance Information Criterion (DIC) approach.

Vaccination strategies

We expand the model structure presented above to incorporate and test the impact of different vaccination strategies in the model (see S1 and S2 Figs for full model structure description). Using the best calibrated model as baseline we introduce four vaccination scenarios, where we combine different levels of vaccine coverage among livestock and humans as well as frequency of campaign roll-out. As follows,

80% of livestock in a single campaign approach

80% of livestock yearly

80% farmers

50% farmers

Each intervention is introduced at year 5 of the simulation, with a linear scale up period of three months. The impact of vaccination is measured as the number of human CCHFV infections averted, and early human deaths averted. To assess efficiency of each approach, we also calculate the ratio of total vaccine courses over human infections averted.

Results

According to our systematic comparison of environmental and epidemiological drivers, a model with Saturation Deficit as a surrogate indicator of tick activity, and an assumption of increased CCHFV reporting capacity, resulted in the best, most parsimonious model fit when assessed with DIC. shows model outputs for the best performing model against calibration targets. Interestingly, the DIC estimate was very close (within 5 units) for most environmental drivers (see ). Only relative humidity displayed a markedly worst fitting scenario. On the other hand, the baseline assumption of increased reporting capacity of human CCHFV was consistently superior to other epidemiological assumptions like the sustained influx of livestock from high endemic areas (). These results suggest a zoonotic endemic transmission that is well captured by the oscillations in the saturation deficit index. In humans, the spill-over would follow the same trend and more importantly, we estimate that only 31% (CrI 95% 28%-33%) of cases would result in symptomatic disease (see ).

Model trajectories against calibration target data.

(A) Simulated age stratified CCHFV prevalence among livestock (green density plot), with the median estimate (white horizontal line), against IgG prevalence data for the same age groups as reported by Mustafa et al [17] from Herat (black square shows the mean and error bars the 95%CI). (B) Posterior density and median estimate of IgG prevalence for the population of farmers and other occupations (density plots pink and blue) against IgG prevalence data from Herat reported. We take the prevalence estimate to match the dates of data collection as reported by Mustafa et al. (C) Stochastic model trajectories (grey lines) for monthly incident CCHFV human cases reported in Herat. In shaded pale grey, the 95% CrI and in solid blue, the median estimate. In black dots, monthly incident cases reported in two separate CCHF outbreaks in Herat: in 2008 as reported by Mofleh et al [15], and 2017–2018 as reported by Niazi et al, and Sahak et al [13,16]. (D) & (E) yearly CCHF cases and deaths reported from Herat, against data (black) as reported by Sahak et al, respectively.

We compare vaccination campaigns directed to animals only, humans only and also combinations of the two and with different campaign frequency. A summary of the overall impact of different vaccination approaches can be found in Overall, vaccination strategies targeted to humans display a much larger impact (on human cases and deaths averted) compared to animal vaccination campaigns. Our results also suggest that human vaccination is a more efficient approach, reflected in less vaccine courses per human case averted: a single campaign for 80% livestock requires about 12 fold the number of vaccine courses to prevent one human case compared to a vaccine campaign reaching 50% of humans (). When we compare campaigns targeted to the overall population vs. farmers it is evident that targeting the high risk groups (farmers) results in higher efficiency (). An increase in the frequency of campaigns targeted to livestock displays a larger epidemiological impact, while resulting in more vaccine courses per case averted over time, compared to a single campaign as seen in . Finally, we simulated different combinations of vaccine efficacy and vaccine coverage for vaccination campaigns in humans (). This analysis shows that for both deaths and infections averted, an intervention targeted to humans yields benefits that are at least one order of magnitude larger compared to the livestock campaign. Importantly, the contour plots show that there is a frontier of high effectiveness that can be reached within a spectrum of combinations of efficacy and coverage.

Population impact of CCHF vaccination strategies.

(A) Cumulative clinical cases averted over 4 years of simulation, for four different scenarios of intervention. (B) the number of vaccine courses required to avert a CCHFV case in humans, estimated as the ratio number of courses over clinical cases averted. The inset window shows a zoom-in for clarity of the two human vaccination interventions. (C), boxplots for the cumulative number of averted clinical cases of CCHFV for the first 4 years of the simulated vaccine period. Inset window shows a zoom-in into the first seven months after vaccination campaigns. In orange, a one-off campaign for vaccinating 50% of humans over a three-month period. In purple, a one-off campaign to vaccinate 80% of livestock over a three-month scale-up period. (D) Sensitivity analysis assessing the effect of the disparity in risk between farmers and other occupations.

Exploration of vaccine efficacy and coverage on incidence and mortality reductions.

(A) A contour for combinations of vaccine efficacy and vaccine coverage among humans. Interventions are introduced as a single campaign approach. White solid lines reflect the frontier of effect measured as CCHFV infections averted. (B) shows the combination efficacy vs human vaccination coverage and effect measured as CCHFV deaths averted. (C) and (D), the effect of different levels of coverage among farmers and other occupations on infections averted and doses per case averted, respectively.

Discussion

Understanding the dynamics and epidemiological drivers of transmission are key to establishing priorities for the research and development roadmap for CCHFV. Here we present for the first time a calibrated mathematical model to simulate the transmission of CCHFV in livestock and spill-over into humans in western Afghanistan.

We find that CCHFV in Herat province has reached an endemic state of transmission within livestock, with a yearly cycle that is well reproduced by the oscillations of the saturation deficit index in this geographical area. This index incorporates air temperature and relative humidity, and indicates that the dry and hot months of summer likely result in periods of high tick activity. As long as tick activity data is mostly absent, this approach will continue to be necessary in the future. We highlight the need to extend this analysis to areas where other environmental drivers could be relevant, or where epidemiological factors result in different types of outbreaks. Spill-over transmission into humans mirrors this seasonal pattern, and although stochastic events explain some of the year-to-year variability, the increasing trend in case reporting appears strongly linked to the increased reporting capacity in the country at the time [13]. Importantly, our results show that the volume of spill-over transmission might be much higher than previously expected: we estimate that 31% (CrI 95% 28% - 33%) of transmission events into humans lead to symptomatic disease and therefore to case reporting. Previous evidence from seroprevalence surveys have estimated higher fractions (ranging 88%-100%) of sub-clinical presentation of CCHFV in humans when contrasted to reported cases [29,30]. Our approach to this wide uncertainty in the existing evidence is to try to infer the fraction symptomatic from the calibration process. Nevertheless, further seroprevalence studies in humans and estimations of the clinical fraction are urgently needed in the field, given the central role this parameter plays in defining the overall burden of disease associated with CCHF and suitable endpoints in future vaccine trials. When estimating the epidemiological impact (in terms of cases and deaths averted) are sensitive to this parameter, as seen in .

Future vaccine campaigns against CCHFV might have the largest population impact when applied to humans instead of animals. In the current study setting, immunisation strategies targeted to farmers (the high risk group) are more efficient as they require less vaccine courses per case averted. It is plausible that other epidemic settings with a more concentrated profile of risk could lead not only to more efficient but more impactful targeted interventions. The latter might require further investigation.

Our results exemplify the challenges posed by animal vaccination: livestock campaigns have a rapid impact, but their effect rapidly wanes as livestock population turnover prevents further accumulation of population immunity. More frequent vaccination campaigns increase the long term impact but not enough to match interventions directed to humans. Another challenge that might arise from a livestock vaccination campaign has to do with the fact that asymptomatic CCHFV infection in livestock might result in poor compliance from farmers and animal owners, as immunisation for innocuous infections will not be a priority. Finally, animal vaccination not only shows in our study to return lower benefits, but it also requires a much larger number of vaccine courses per human case averted.

The current study is restricted to one location, and our assessment of environmental drivers could yield different results in different ecological and climatic settings. A natural progression of this analysis would be to assess this modelling approach in other countries where endemic CCHFV transmission is suspected or established. Another limitation directly related to this is the lack of data on tick activity and tick abundance for this setting. Such data are rare not only in this context but in any setting around the world. Furthermore, by not incorporating an explicit tick-vertebrate or tick-human mechanism in the model there is a possibility that other factors affecting tick populations or tick activity could result in different epidemic trajectories. However, a strength of our analysis is the solution we provide by systematically testing environmental surrogates for tick activity using climatic factors that are well known predictors of tick activity [31]. There is a clear need to collect and analyse tick activity in addition to wildlife host data to better understand the drivers of CCHFV transmission [32].

Our approach focuses on the transmission into humans from human-animal contact, and we ignore human-to-human transmission. Previous evidence shows that human-to-human transmission is plausible and nosocomial transmission has been reported before [33,34]. However, we expect this aspect of transmission to contribute marginally to the annual reported trends of CCHFV cases in Afghanistan as most infections arise outside the hospital environment and are linked to animal handling activities [15,16].

Our assumption on transmission is also central for interpreting the impact of vaccination campaigns, since the absolute epidemiological impact is necessarily limited by the population size of the targeted human group. For this reason, a measure of efficiency like doses per case averted might be a better indicator of intervention performance.

In this work we also ignore transmission cycles in wildlife. This can be important for maintaining more stable levels of endemicity, but the absence of data prevents us from designing a more complex transmission network for Herat.

In conclusion, CCHFV is likely to be endemic in western Afghanistan, with a seasonal pattern which is robustly predicted by climatic factors which we explore in this work. The increasing number of human cases reported in Herat are most likely explained by increasing trends in reporting capacity in the country, and more importantly, these cases are reflecting only a fraction of the overall volume of human infection. Vaccination campaigns in humans are more impactful and efficient in the medium and long term compared to livestock vaccination. Finally, targeting campaigns to groups with an increased risk of infection, like farmers, are the most efficient strategies in our assessment and should be a key component of future vaccine implementation roadmaps for CCHFV.

Model structure for the transmission of CCHFV amongst livestock.

The structure shows the compartments and state transitions for livestock. Births occur at a rate equal to the mortality rate to maintain a population at equilibrium. Mortality rates are estimated to achieve a known livestock age-distributed population (see S2 Text). Offspring from prevalent CCHFV animals acquire transient immunity at birth through first colostrum (Ri). This immunity lasts for an assumed average period of 6 months. After this period, livestock move to the susceptible stage (S1). Susceptible livestock (Sa) acquire CCHFV with a force of infection λ that leads to an infectious period (Ia) with mean duration D, expressed as inverse time rate D. We assume that livestock lose immunity at a rate D Vaccination is implemented by recruiting susceptible animals at a rate ν(t). The effective number of immunised livestock is also defined by the efficacy of the vaccine (κ). At first, vaccination does not confer immunity V, which is only acquired after a period of length D In this structure, subscript a points to the age category within the age structure, over which transitions occurs as depicted with the shaded grey structure in the background

(TIF)

Click here for additional data file.

Model structure for the spillover transmission of CCHFV and disease progression in humans.

Humans are born at a rate reflecting the life expectancy in Afghanistan (keeping population size constant in the absence of infections), and split into the two human categories considered in this model. Namely farmers (the high-risk group) and other occupations. This distribution is taken from previous USAID surveys in the country (see parameters Table 1 in the main text). The categorisation by occupation in the model is reflected in this structure and in the mathematical equations using subscript k (0 = farmer; 1 = others). Infection is acquired in humans with force of infection λ, with differential risk k. Infection is followed by a latent period with mean duration D that leads to an infectious period . This infectious period can lead to either recovery or death. Death from CCHFV in humans is described in the model as the competing hazard μ that summarise the case fatality ratio for CCHFV. We assume waning immunity in humans that leads back to susceptible stage at a rate D. Vaccination occurs at rate (t), differential by occupation. Effective number of immunised people is finally defined by the efficacy of the vaccine (κ). Vaccine protection occurs after vaccination after a mean period D.

Table 1

Model parameters.

Parameter description	Notation	Input Values/Estimated*	Source
Natural history of disease
Livestock
Duration of infectiousness in livestock	D iL	7 days	Gonzalez et al., 1998[20]
Duration of colostrum acquired immunity (months)	D aL	8.3 (CrI 95% 2–10)	Estimated
Mean time to loss of immunity in adult livestock (months)	D mL	52 (CrI 95% 46–76)	Estimated
Proportion of livestock immune at time 0 by age¥ group a	Ra(t)	R_a(t)={0.29fora=10.48fora=20.8fora=30.87fora=40.87fora=5	Barthel et al., 2014[21]
Humans
Duration of latent period in humans	D lH	4 days	Bente et al., 2013[22]
Duration of infectiousness in humans	D iH	9 days	Fillâtre et al., 2019[23]
Duration of immunity in humans	D mH	3650 days	Assumption
Fraction of human infection resulting in a clinical case	ϕ	0.31 (CrI 95% 0.28–0.33)	Estimated
Proportion of farmers immune at time 0	pF	0.1333	Mustafa et al., 2011[17]
Proportion of others immune at time 0	pO	0.0469	Mustafa et al., 2011[17]
Case fatality rate of CCHF	CFR cchfv	0.33	Niazi et al., 2019(16)
Demographics
Livestock population size	NL	15,193	FAO 2008 [24]
Livestock ageing factor (1/months)	δ	1/12	Assumption
Livestock monthly death rate	μ	μ_a={0.0761fora=10.0743fora=20.0746fora=30.0744fora=40.0747fora=5	See Fig A in S2 Text
Population size—Farmers	NF	7,614	USAID 2008
Population size—Other occupations	NO	17,768	USAID 2008
Life expectancy—humans	LH	61.5 years	World bank 2008–2014[25]
Monthly birth rate humans	bH	1/ (12*61.5)	Assumption
Monthly birth rate in livestock	bL	μ	Assumption
Viral transmission parameters
Between livestock transmission temperature dependent	A	0.33 (CrI 95% 0.2–0.4)	Estimated
Transmission rate from livestock to farmers	βF	0.28 (CrI 95% 0.15–0.34)	Estimated
Other occupations relative transmission factor(relative to farmers)	O	0.3 (CrI 95% 0.1–0.5)	Estimated
Transmission rate from livestock to other occupations	βo	Oβ F	Assumption
Vaccination parameters
Vaccine efficacy	κ	90%	Assumption
Time to vaccine protection	D pV	2 weeks	Assumption

*Estimated values represent the posterior mean and 95% CrI for the best most parsimonious model, i.e., saturation deficit obtained during calibration (see section S3 Text for calibration details).

¥ Livestock age stratification groups where a = 1 reflects 0 to 12 months; a = 2 for 13 to 24 months; a = 3 for 25 to 36 months; a = 4 for 37 to 48 months, a = 5 for 48 months and older

(TIF)

Click here for additional data file.

Model trajectories against calibration target data for a model with lifelong immunity protection among livestock.

Panel A shows the age stratified simulated CCHFV IgG prevalence among livestock (green density plot), with the median estimate (white horizontal line), against IgG prevalence data for the same age groups as reported by Mustafa et al [14] from Herat (black square shows the mean and error bars the 95%CI). Panel B shows the posterior density and median estimate of IgG prevalence for the population of farmers and other occupations (density plots pink and blue) against IgG prevalence data from Herat reported. We take the prevalence estimate to match the dates of data collection as reported by Mustafa et al. Panel C shows stochastic model trajectories (grey lines) for monthly incident CCHFV human cases reported in Herat. In shaded pale grey, the 95% CrI and in solid blue, the median estimate. In black dots, monthly incident cases reported in two separate CCHF outbreaks in Herat: in 2008 as reported by Mofleh et al [16], and 2017–2018 as reported by Niazi et al, and Sahak et al [15,17]. In Panels D and E, yearly CCHF cases and deaths reported from Herat, against data (black) as reported by Sahak et al.

(TIF)

Click here for additional data file.

Deviance information Criterion DIC values for different modelling assumptions on environmental driver, case reporting assumption, and duration of livestock acquired immunity.

As mentioned in the main text, saturation deficit with baseline assumptions about reporting produces the lowest DIC (best model fit). Importantly, a model with lifelong immunity among livestock shows the worst performance.

(TIF)

Click here for additional data file.

Transmission dynamics of CCHF in Herat, Afghanistan 2008–2018 In panel A, simulated trajectories of monthly CCHF incidence in a spectrum from reported clinical cases (blue shade), to all clinical cases (green) and all cases (red) including symptomatic/subclinical cases. The shaded area shows the 95% CrI. In Panel B, the simulated effective reproduction number for CCHFV in livestock. These are results for the final selected model, i.e., “saturation deficit driver” model.

(TIF)

Click here for additional data file.

Effect of clinical fraction on the epidemiological impact of human vaccination.

Increase in vaccination coverage among humans results as expected in increased epidemiological impact, in infections (panel A), clinical cases (panel B) and deaths (panel C). We take extreme values for the clinical fraction parameter to show that a scenario where only 5% of infections are symptomatic (blue boxplots) results in a drastic decline in epidemiological impact in cases and deaths when compared against a highly symptomatic scenario (green boxplots). The baseline calibrated model is shown in orange, for which the mean clinical fraction is 31%.

(TIF)

Click here for additional data file.

Model equations; Table A in S1 Text. Age distributed CCHFV prevalence among livestock; Fig A in S1 Text. Polynomial model prediction model of Saturation deficit on Air temperature.

(DOCX)

Click here for additional data file.

Livestock demographic model; Fig A in S2 Text. Age distribution in cattle into 5 age yearly groups.

(DOCX)

Click here for additional data file.

Model calibration; Table A in S3 Text. Calibration target datasets for CCHFV in Herat, Afghanistan; Fig A in S3 Text. MCMC Trace plots; Fig B in S3 Text. Density plots; Fig C in S3 Text. Gelman-Rubin diagnostic.

(DOCX)

Click here for additional data file.

Environmental drivers.

(DOCX)

Click here for additional data file.

21 Mar 2022

Dear Dr Vesga,

Thank you very much for submitting your manuscript "Transmission dynamics and vaccination strategies for Crimean-Congo haemorrhagic fever virus in Afghanistan: a modelling study" for consideration at PLOS Neglected Tropical Diseases. As with all papers reviewed by the journal, your manuscript was reviewed by members of the editorial board and by several independent reviewers. The reviewers appreciated the attention to an important topic. Based on the reviews, we are likely to accept this manuscript for publication, providing that you modify the manuscript according to the review recommendations.

The reviewers all agreed that this manuscript is valuable and only needs a few minor clarifications/corrections before publication. Reviewer 1 commented on the source of some assumptions made in the model, so please be sure the assumptions are well-explained and well-cited. The remaining comments from reviewers are primarily editorial changes. Please consider them in your revision.

Please prepare and submit your revised manuscript within 30 days. If you anticipate any delay, please let us know the expected resubmission date by replying to this email.

When you are ready to resubmit, please upload the following:

[1] A letter containing a detailed list of your responses to all review comments, and a description of the changes you have made in the manuscript.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out

[2] Two versions of the revised manuscript: one with either highlights or tracked changes denoting where the text has been changed; the other a clean version (uploaded as the manuscript file).

Important additional instructions are given below your reviewer comments.

Thank you again for your submission to our journal. We hope that our editorial process has been constructive so far, and we welcome your feedback at any time. Please don't hesitate to contact us if you have any questions or comments.

Sincerely,

Brianna R Beechler, Ph.D., DVM

Associate Editor

PLOS Neglected Tropical Diseases

Jeremy V. Camp, PhD

Deputy Editor

PLOS Neglected Tropical Diseases

***********************

The reviewers all agreed that this manuscript is valuable and only needs a few minor clarifications/corrections before publication. Reviewer 1 commented on the source of some assumptions made in the model, so please be sure the assumptions are well -explained and well-cited. The remaining comments from reviewers are primarily editorial changes. Please consider them in your revision.

Reviewer's Responses to Questions

Key Review Criteria Required for Acceptance?

As you describe the new analyses required for acceptance, please consider the following:

Methods

-Are the objectives of the study clearly articulated with a clear testable hypothesis stated?

-Is the study design appropriate to address the stated objectives?

-Is the population clearly described and appropriate for the hypothesis being tested?

-Is the sample size sufficient to ensure adequate power to address the hypothesis being tested?

-Were correct statistical analysis used to support conclusions?

-Are there concerns about ethical or regulatory requirements being met?

Reviewer #1: The objectives of the study were clearly stated in the author summary and introduction of the manuscript. The authors state that the study is to develop an analytical model looking at different ecological and epidemiological variables related to CCHF to shed light on its transmission dynamics in humans and what effect vaccination of different populations may have on these dynamics.

The study design was a mathematical model analyzing various inputs related to ecological and epidemiological factors for CCHF virus and was appropriate given the stated objectives. The methodology and inputs were appropriate and reasonable given the study questions being addressed. There was a comprehensive review of various inputs for all aspects and stages of the livestock, human and vector components of CCHF transmission cycles. There were a number of inputs and values that were “estimated” and it was not always clear how these inputs and values were determined.

A few clarifications related to the analysis and variables and values used in the model analyses:

• It was not clear if the authors accounted for varying contact rates with livestock between farmers and the general population. The calibrated model shows the proportion of contact for other occupational groups was 0.3 and transmission from livestock to farmers was 0.28. Was this assumption based on farmers having higher pre-existing CCHF immunity?

• The authors estimate waning CCHF immunity on livestock after 5 years. What was this based on? And even if waning immunity, by what factor would this reduce protection against re-infection.

• The model assumes these animals after 5 years are able to be re-infected, but the general consensus for CCHF is infection confers lifelong immunity for both animals and humans.

• The human immunity is also estimated at 10 years, and we presume lifelong immunity, or at least protective immunity even if waning measurable antibody titers.

• The authors estimate the fraction of human CCHF infections that result in clinical cases was .31. Based on first-hand experience and study this seems very low and could be an underestimate.

• These assumptions may impact the model outputs if not validated or evidence is available for documented re-infection of humans and/or animals

• Authors state ticks bury in soil once temp goes beyond 30 degrees. What is the shape of the function for declining transmission? Is it linear? Exponential? What assumptions are made? Was this determined by model calibration?

No ethical concerns were noted.

Reviewer #2: Yes the objectives are clear on the design appropriate to achieve the objectives. Also correct statistics has been done but this would be checked by a statistician or someone who is a stronger mathematical background to make sure that today models are aware formulated and calibrated

Reviewer #3: None.

--------------------

Results

-Does the analysis presented match the analysis plan?

-Are the results clearly and completely presented?

-Are the figures (Tables, Images) of sufficient quality for clarity?

Reviewer #1: The analysis plan described in the manuscript of the mathematical model was appropriate and straight forward. All analysis and results were clearly described and the relevant findings were highlighted. The results appear to be reasonable and as expected based on the assumptions, inputs and general epidemiology and impact of vaccination on the proposed target populations.

The tables and figures accurately represent the data presented and are thorough and detailed.

Reviewer #2: The results are represented and well supported by the supplementary material that gives the details of some of the analysis and their their results

Reviewer #3: The results are clearly and completely presented. The figures and tables are of sufficient quality.

--------------------

Conclusions

-Are the conclusions supported by the data presented?

-Are the limitations of analysis clearly described?

-Do the authors discuss how these data can be helpful to advance our understanding of the topic under study?

-Is public health relevance addressed?

Reviewer #1: The overall discussion and conclusions stated by the authors agrees with the data and analysis presented. In general the authors discuss their model findings and the impact of vaccination on the estimated transmission of CCHF. A few items requiring further clarification:

• The first point of discussion highlights the lack of tick vector data in which to calibrate and use for the model. It is not stated in the manuscript what the assumption of percent of total ticks are infected with CCHF and how this impacts the potential infectivity of cattle/human populations. From practical experience tick infectivity could be low, but the documented seropositivity in livestock is often very high. Therefore the estimate for livestock-livestock CCHF transmission could be happening at a higher rate than estimated in the model, although the estimates used could offset each other and therefore not have any impact on this model or effect any of the resulting outputs.

• The authors state again that an estimated 31% of spillover cases become symptomatic and therefore detectible by surveillance. As stated above, this seems too low of an estimate. The two references (29, 30) showing that the potential for asymptomatic infection for CCHF could be higher. I believe the authors may have misinterpreted these references and greatly overestimated the number of potential asymptomatic infections. One of the references use a serological assay known for its overestimation of the true IgG seropositivity. The second reference is a validation study of IgG positive CCHF cases from three independent serosurveys where the IgG seropositivity that could be attributable to asymptomatic CCHF as less than 5%.

Generally, the effects of vaccination on the targeted populations and the resulting reduction in estimated CCHF human cases is as would be expected given the analysis. The only change to the results could be the absolute number calculated from the model if some of the assumptions were to be adjusted up or down based on more real-world estimates.

Reviewer #2: The study is a good public health importance because it touches and emerging disease that has not been given attention by the public health bodies. A disease that causes mortality and mobility and no vaccine is in place yet to be used especially in high-risk groups such as farmers and Herdsmen. However more work would be needed to implement some of these recommendations in the article. One research or one publication is not enough until some of the work is done in the different countries where CCHF is prevalent so as to give a comprehensive picture. The authors could edit their recommendations and conclusions accordingly.

Reviewer #3: The conclusions are well supported by model results. The limitations, particularly around the large gap in information for CCHF, are well described. The results are contextualized well and the public-health relevance is well addressed.

--------------------

Editorial and Data Presentation Modifications?

Use this section for editorial suggestions as well as relatively minor modifications of existing data that would enhance clarity. If the only modifications needed are minor and/or editorial, you may wish to recommend “Minor Revision” or “Accept”.

Reviewer #1: none

Reviewer #2: The manuscript could receive another read and make sure that the English drama is reading well especially in the abstract. For example I do not know the use of the word"course "as used in this abstract or are they meaning "Cases".

Reviewer #3: Main text

------------

P 7: The formula at the end of Figure 1 caption should have N_L instead of N, like the equation on line 136.

P 7: Lines 120—125 should continue the caption of Figure 1.

P 10: What is the value of the monthly birth rate in livestock?

P 10: The notation for transmission from livestock to other occupational groups should be \\beta_O.

P 10: What is the value of the time to vaccine protection?

P 10: Lines 176–178 are an odd place for the ethics statement. Perhaps move it elsewhere.

P 17: On line 251, "panel B of Figure 3B" should be "Figure 3B" in parallel with "Figure 3D" on line 258.

Supplement

----------

P 4–6: What are the initial conditions for the livestock model? How long is period of time is simulated?

P 4: The second line of Eq. 1 is not rendered correctly.

P 4: Comparing Eq. 2 to Figure S1, the term for births from recovered livestock (R_a(t)) is missing.

P 5: In Eq. 3, the "L" following lambda should be a subscript.

P 5: In Eq 7, for Z_{i,j}, the lower left term Z_{5,5} should be 0, not -1, as the animals 5th age group do not age out, only leave through mortality (mu_a).

P 6–8: What is the time step used in the human model? What are the initial conditions for the human models? How long is period of time is simulated?

P 8 & 9: I don't understand how saturation deficit, air temperature, and tick activity are related from the section and Figure S3. What are the black dots in the figure? Please explain further.

P 10: I found "with the last categories being those 4+ years" to be confusing. I recommend "except the last age group includes all those aged 4+ years" or similar.

P 10: "It has been observed that Hyalomma spp. similarly to other tick species, are strongly driven in their reproduction cycles and also feeding activity by different climatic and land composition variables." Please provide citations at the end of this sentence.

P 11: "prevalent I" should be "prevalence".

P 12: "As a rule of thumb, values below 1.1 are typically considered to indicate convergence." Please provide a citation, perhaps to Gelman's book.

P 12: You refer to the test statistic (R with two dots) as the "scale reduction factor" then the "Potential scale reduction factor (PSRF)" in the same paragraph, with the latter also used in Figure S7. Please either use "potential" throughout or explain why "potential" is added later.

P 12: "shoes" should be "shows".

--------------------

Summary and General Comments

Use this section to provide overall comments, discuss strengths/weaknesses of the study, novelty, significance, general execution and scholarship. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. If requesting major revision, please articulate the new experiments that are needed.

Reviewer #1: Overall, this manuscript is a good first analysis of the impact vaccination could have on human CCHF infections given the transmission dynamics in a specific geographic location in Afghanistan. It may not be as generalizable given the dynamics of CCHF varies so widely across different geographic locations and ecological niches. Some limitations the authors properly address, but lack of location specific variables/data points could bias their analysis if taken from locations distant from the study location (e.g CCHF data from Turkey) to help validate or calibrate the model.

Reviewer #2: This manuscript if published would you give a body of scientific knowledge on how to model infectious diseases especially vector- born infectious diseases. So it's of good public health importance and I would recommend it gets published. It also presents a challenge which is mentioned as a limitation for not getting tick activity data and using environmental data as a proxy which is a challenge that we see in most of the modeling studies involving vector dynamics. The authors could also mention the practicability of the recommendations in the study country. How easy is it to vaccinate cattle.animals, farmers or high risk in study country. If these recommendations taken on; are they practical.

Reviewer #3: The manuscript presents a novel modeling study of Crimea–Congo hemorrhagic fever (CCHF) in Afghanistan that provides a solid footing for important and relevant findings. The model used has an appropriate balance of simplicity and use of available data, which is particularly important due to the complex transmission cycle of CCHF and complex life cycle of its tick vectors. Key missing information was estimated using Bayesian methods and model selection, which is appropriate.

Model selection showed that saturation deficit was the most important environmental driver of infections (through its effect on ticks) and that recent increases in incidence were best explained by improved reporting, rather than an increase in livestock imports. The latter seems like an important finding, although it was not heavily emphasized.

Most importantly, the work shows that, for a hypothetical vaccine, vaccinating humans is much more effective at reducing infections, clinical cases, and deaths in humans than vaccinating livestock, and vaccinating farmers, who at high risk due to contact with livestock, is particularly efficient per dose of vaccine. These are important and relevant findings for mitigating the impact of this disease.

--------------------

PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: Yes: Luke Nyakarahuka

Reviewer #3: Yes: Jan Medlock

Figure Files:

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org.

Data Requirements:

Please note that, as a condition of publication, PLOS' data policy requires that you make available all data used to draw the conclusions outlined in your manuscript. Data must be deposited in an appropriate repository, included within the body of the manuscript, or uploaded as supporting information. This includes all numerical values that were used to generate graphs, histograms etc.. For an example see here: http://www.plosbiology.org/article/info%3Adoi%2F10.1371%2Fjournal.pbio.1001908#s5.

Reproducibility:

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

References

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article's retracted status in the References list and also include a citation and full reference for the retraction notice.

1 Apr 2022

Submitted filename: Reviewers_comments_final.docx

Click here for additional data file.

29 Apr 2022

Dear Dr Vesga,

We are pleased to inform you that your manuscript 'Transmission dynamics and vaccination strategies for Crimean-Congo haemorrhagic fever virus in Afghanistan: a modelling study' has been provisionally accepted for publication in PLOS Neglected Tropical Diseases.

Before your manuscript can be formally accepted you will need to complete some formatting changes, which you will receive in a follow up email. A member of our team will be in touch with a set of requests.

Please note that your manuscript will not be scheduled for publication until you have made the required changes, so a swift response is appreciated.

IMPORTANT: The editorial review process is now complete. PLOS will only permit corrections to spelling, formatting or significant scientific errors from this point onwards. Requests for major changes, or any which affect the scientific understanding of your work, will cause delays to the publication date of your manuscript.

Should you, your institution's press office or the journal office choose to press release your paper, you will automatically be opted out of early publication. We ask that you notify us now if you or your institution is planning to press release the article. All press must be co-ordinated with PLOS.

Thank you again for supporting Open Access publishing; we are looking forward to publishing your work in PLOS Neglected Tropical Diseases.

Best regards,

Brianna R Beechler, Ph.D., DVM

Associate Editor

PLOS Neglected Tropical Diseases

Jeremy V. Camp, PhD

Deputy Editor

PLOS Neglected Tropical Diseases

***********************************************************

Editorial Note: The reviewers are happy with your revisions. Reviewer 3 noted that a few equations did not render correctly in the most recently submitted version so be sure to check the proofs carefully to ensure equations are showing correctly.

Reviewer's Responses to Questions

Key Review Criteria Required for Acceptance?

As you describe the new analyses required for acceptance, please consider the following:

Methods

-Are the objectives of the study clearly articulated with a clear testable hypothesis stated?

-Is the study design appropriate to address the stated objectives?

-Is the population clearly described and appropriate for the hypothesis being tested?

-Is the sample size sufficient to ensure adequate power to address the hypothesis being tested?

-Were correct statistical analysis used to support conclusions?

-Are there concerns about ethical or regulatory requirements being met?

Reviewer #1: Methods have been clarified. Revisions are sufficient for acceptance

Reviewer #3: (No Response)

**********

Results

-Does the analysis presented match the analysis plan?

-Are the results clearly and completely presented?

-Are the figures (Tables, Images) of sufficient quality for clarity?

Reviewer #1: Questions related to results have been addressed by reviewers. Although not completely able to be addressed, the limitations of this have also been highlighted and readers are aware.

Reviewer #3: (No Response)

**********

Conclusions

-Are the conclusions supported by the data presented?

-Are the limitations of analysis clearly described?

-Do the authors discuss how these data can be helpful to advance our understanding of the topic under study?

-Is public health relevance addressed?

Reviewer #1: Conclusions are sufficient for accceptance

Reviewer #3: (No Response)

**********

Editorial and Data Presentation Modifications?

Use this section for editorial suggestions as well as relatively minor modifications of existing data that would enhance clarity. If the only modifications needed are minor and/or editorial, you may wish to recommend “Minor Revision” or “Accept”.

Reviewer #1: (No Response)

Reviewer #3: I appreciate the authors' attention to the minor issues I raised in the previous round of reviews. Except as noted below, the authors resolved the issues I raised.

Many of the equations in the supplement, which was only available as DOCX, did not render at all for me. I believe this was some software compatibility problem. Thus, I was unable to confirm the issues from my first review that are listed below were resolved.

Supplement

----------

P 4: The second line of Eq. 1 is not rendered correctly.

P 4: Comparing Eq. 2 to Figure S1, the term for births from recovered livestock (R_a(t)) is missing.

P 5: In Eq. 3, the "L" following lambda should be a subscript.

P 5: In Eq 7, for Z_{i,j}, the lower left term Z_{5,5} should be 0, not -1, as the animals 5th age group do not age out, only leave through mortality (mu_a).

**********

Summary and General Comments

Use this section to provide overall comments, discuss strengths/weaknesses of the study, novelty, significance, general execution and scholarship. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. If requesting major revision, please articulate the new experiments that are needed.

Reviewer #1: None. Nice first paper to assess the impact of vaccination on CCHF

Reviewer #3: (No Response)

**********

PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #3: Yes: Jan Medlock

18 May 2022

Dear Dr Vesga,

We are delighted to inform you that your manuscript, "Transmission dynamics and vaccination strategies for Crimean-Congo haemorrhagic fever virus in Afghanistan: a modelling study," has been formally accepted for publication in PLOS Neglected Tropical Diseases.

We have now passed your article onto the PLOS Production Department who will complete the rest of the publication process. All authors will receive a confirmation email upon publication.

The corresponding author will soon be receiving a typeset proof for review, to ensure errors have not been introduced during production. Please review the PDF proof of your manuscript carefully, as this is the last chance to correct any scientific or type-setting errors. Please note that major changes, or those which affect the scientific understanding of the work, will likely cause delays to the publication date of your manuscript. Note: Proofs for Front Matter articles (Editorial, Viewpoint, Symposium, Review, etc...) are generated on a different schedule and may not be made available as quickly.

Soon after your final files are uploaded, the early version of your manuscript will be published online unless you opted out of this process. The date of the early version will be your article's publication date. The final article will be published to the same URL, and all versions of the paper will be accessible to readers.

Thank you again for supporting open-access publishing; we are looking forward to publishing your work in PLOS Neglected Tropical Diseases.

Best regards,

Shaden Kamhawi

co-Editor-in-Chief

PLOS Neglected Tropical Diseases

Paul Brindley

co-Editor-in-Chief

PLOS Neglected Tropical Diseases

Table 2

Environmental drivers as surrogate markers of tick activity.

Environmental indicator	Description	How we modelled it	Source of data
Soil temperature (ST)(Celsius)	Temperature of the soil in the first layer (0–7 cm) taken at 10:00 AM	Vector of monthly average from April 2008 to January 2019. We assume a tick activity range between 12°C and 30°C.	ERA5 atmospheric variables, centred in a polygon in Herat (ECMWF and Copernicus[26] [27])
Relative humidity (RH)	It is a measure of vapor content in the air.	Vector of monthly average from April 2008 to January 2019. As ticks prefer dry hot weather, we use the complement (1-RH) to indicate increase in tick activity	Constructed from air temperature (T), dew point temperature (Td), and surface pressure from ERA5 (ECMWF and Copernicus[26] [27])
Saturation deficit (SD)	A measure of the drying power of the air. It accounts both for air temperature, vapor pressure and relative humidity.	Vector of monthly average from April 2008 to January 2019. Given that SD includes temperature, we use a simple regression model to find the SD range of tick activity matching the ST range.	ERA5 atmospheric variables centred in a polygon in Herat (ECMWF and Copernicus[26] [27])
Normalised difference vegetation index (NDVI)	Combines satellites signals to estimate the density of green on an area of land. It indicates a combination of rainfall, and land change.	Vector of monthly average from April 2008 to January 2019.	NASA, EarthData (MODIS/VIIRS subsets) for Herat [28]

Table 3

Model comparison using DIC.

Step 1: Selection of environmental driver assumption	DIC
Saturation deficit	65.21
NDVI	69.1
Soil temperature	70.05
Relative humidity	86.81
Step 2: Selection of Saturation deficit + Epidemiological assumption
Improved reporting	65.21
Increase influx of livestock (stable reporting)	77.64
Increase influx of livestock and improved reporting over time	79.05

Table 4

Epidemiological impact of modelled vaccination strategies.

CCHFV Infections and early deaths averted, and number of vaccine courses per clinical case averted, according to the four vaccination scenarios, cumulatively over the period April 2014 to Dec 2018.

Vaccination scenario	Human CCHF infections averted (CrI 95%)	Human clinical CCHF cases averted (CrI 95%)	Human CCHF deaths averted (CrI 95%)	Total vaccine courses* (CrI 95%)	Vaccine courses per clinical case averted(CrI 95%)
80% of livestock in a single campaign	105 (38–207)	31 (10–65)	10 (2–22)	12,578 (8,857–27,141)	431 (162–1,438)
80% of livestock yearly	318 (117–632)	94 (30–198)	31 (10–66)	108,948 (73,236–260,514)	1,243 (465–4,389)
50% humans in a single campaign	902 (326–1832)	266 (88–568)	87 (28–185)	9,533 (8,294–10,241)	35 (16–107)
80% farmers in a single campaign	686 (270–1039)	191 (59–490)	63 (19–164)	3,700 (3,060–4,242)	19 (7–62)

* Cumulative vaccine courses over simulation period