The Reproduction Number of Swine Viral Respiratory Diseases: A Systematic Review

Simple Summary Diseases in the swine industry can cause major economic and health issues. This review examines R0 estimates, which measure how efficiently respiratory diseases spread among pigs and compares transmission risks within and between farms. Seven diseases were studied: Aujeszky’s disease, PPRSV, Circovirus, Influenza A, Encephalomyocarditis Virus, Classical Swine Fever, and African Swine Fever. Data from research articles showed varying R0 values, with higher transmission rates within herds for some diseases. Effective disease control requires prompt detection, response to outbreaks, and continuous monitoring to reduce its impact on the swine industry. Abstract Diseases in the swine industry can cause significant economic and health impacts. This review examines R0 estimates for respiratory diseases in pigs, assessing variations and comparing transmission risks within and between farms. A literature search of three databases aggregated peer-reviewed research articles on swine viral respiratory diseases’ R0 values. The study focused on seven diseases: Aujeszky’s disease (AD), Porcine Reproductive and Respiratory Syndrome Virus (PRRSV), Circovirus, Influenza A (IA), Encephalomyocarditis Virus (EV), Classical Swine Fever (CSF), and African Swine Fever (ASF). R0 values were estimated for transmission within and between herds/farms using various methods, from complex mathematical models to simple calculations. Data primarily came from disease surveillance and laboratory experiments. The median R0 for within-herd and between-herd transmission was 10 and 3.31 for AD, 2.78 and 1.14 for PRRSV, 5.9 and 0.89 for Circovirus, 1.75 and 1.6 for CSF, and 3.94 and 3.15 for ASF. For IA and EV, only within-herd R0 values were estimated at 8.65 and 1.3, respectively. Diseases with high R0 values highlight the need for prompt detection and response to outbreaks. Continuous monitoring and evaluation of pathogen transmissibility are crucial for enhancing disease surveillance and reducing the impact of livestock diseases.


Introduction
Pork is a leading source of protein consumption worldwide, making up a third of all meat production globally [1].Therefore, diseases in the swine industry are likely to have significant economic and indirect health impacts in both pork-producing and porkconsuming communities.A disease is considered economically important if it leads to death, wasting, decreased efficiency of gain, reduced rate of gain, immune suppression/increased susceptibility to other disease conditions, or infertility in livestock species used as food [2,3].Respiratory diseases are one of the most important causes of economic losses in pork production [2][3][4].
In this study, we review seven viral diseases with various respiratory clinical presentations and transmission routes.Respiratory symptoms include but are not limited to Vet.Sci.2024, 11, 300 2 of 25 coughing, sneezing, shortness of breath (dyspnea), and inability to grow up to production standards [1,5,6].Aujeszky's disease, caused by suid herpesvirus 1, is a respiratory syndromic disease that can be spread by various transmission routes, including direct contact, airborne, and fomite transmission [7].While the disease has been mostly eradicated in many countries in Europe and North America, it remains endemic in areas like south-eastern Europe and Latin America [8].Circoviruses, such as Porcine Circovirus 2 and 3 (PCV2 and PCV3), are emerging pathogens in the pork industry that impact multiple organ systems, including the respiratory system [9,10].Studies indicate that PCV2 has been detected in air samples, and evidence indicates that there is short-distance transmission [11,12].Encephalomyocarditis in pigs is best categorized as a respiratory syndrome, as transmission occurs primarily through contaminated feed and water [13,14].African Swine Fever (ASF) was believed to be primarily transmitted through secretion and only has a clinical respiratory manifestation, but recent research has evidence of aerosol transmission through short distances [15,16].Like ASF, Classical Swine Fever has evidence of short-distance aerosol transmission especially high-virulence strains [17][18][19].Both Influenza A Virus (IAV) and Porcine Reproductive and Respiratory Syndrome Virus (PRRSV) spread through aerosol transmission, and the disease impacts gain, often leading to death in piglets [20][21][22][23][24].While acute and dramatic clinical signs may resolve relatively quickly, this does not mean that their subclinical effects and costs are self-limiting or resolve quickly.Because pork makes up a significant portion of global meat production, research on swine diseases has been primarily driven by the needs of the food industry [1].
The basic reproductive number (R 0 ) is the expected number of secondary infections generated by an infectious individual in an otherwise susceptible population.When R 0 is less than one, the disease outbreak dies off as each infectious individual produces less than one new infection on average.Conversely, when R 0 is above the threshold value, each infection generates multiple new infections, allowing the disease to spread through a population.This epidemiological metric has been widely used to inform the risk and potential magnitude of a disease outbreak, the critical herd size required for pathogen persistence, and the efficacy or magnitude of control efforts [1,25].This review summarizes R 0 estimates of respiratory diseases in pigs, as reported in the literature, evaluates variations in reported R 0 values, and compares estimates of disease transmission risk within and between pig farms.

Materials and Methods
A literature search was conducted in the PubMed, Web of Science, and EMBASE (OVID) databases to aggregate peer-reviewed original research articles on R 0 of pig viral respiratory diseases.The PRISMA guidelines were utilized to identify, screen, and determine the eligibility of the articles.A list of diseases was identified by searching the Merck Veterinary Manual and CABI Compendium by combining the terms "respiratory disease" and "pigs" [6,26].Search terms were aggregated into disease and reproductive number categories.The terms "swine", "pig", and "porcine" were used to narrow the study populations.Specific viral diseases and general terminology for respiratory diseases were combined to make a disease query.The full list of search queries is provided in Table 1.All citations were imported into EndNote 21 (EndNote Version 21.2, Clarivate, Philadelphia, PA, USA).
The search returned a combination of modeling exercises, laboratory experiments, and observed outbreaks.The exclusion criteria were as follows: bacterial diseases, studies that only reported the reproductive number for intervention strategies (i.e., vaccination), research not classified as original research, and articles not published in English.To be included, the paper had research that applied to domestic pig production.From the remaining studies, information on herd classification, modeling approach, and method of calculating R 0 was extracted.The herd classifications were within-and between-herd to specify whether a paper-modeled disease spread within a single farm population or across multiple farms.A brief statement about the type of model used in each paper was created from their method section.

Category Terms
Mathematical "Reproduction number" OR "reproductive number" OR "reproductive rate" OR "reproductive ratio" OR "reproduction ratio" OR "R0" OR "force of infection" OR "secondary infection" OR "secondary case" OR "attack rate" OR "attack ratio" OR "population dynamics" OR "infection dynamics" OR "transmission rate" or "R0 Structural Parameters" Animal livestock OR farm * OR pig OR porcine OR swine Disease "Respiratory virus" OR "Respiratory disease" OR paramyxovirus OR herpesvirus OR alphaherpesvirus OR "adenovirus" OR "viral pneumonia" OR "porcine respiratory disease" OR "porcine infectious disease" OR "porcine coronavirus" OR "porcine respirovirus 1" OR "swine orthopneuovirus" OR "swine influenza" OR "Porcine reproductive and respiratory virus" OR PRRSV OR "syndrome PRRSV Virus" OR "swine influenza" OR SIV OR "African Swine Fever" OR ASF OR ASFV OR "suid herpesvirus 1" OR SHV1 OR "Porcine Epidemic Abortion and Respiratory Syndrome" OR "classical swine fever" OR "CSF" OR "H1N1"OR "H3N2" OR "H1N2"OR "PRCV" OR "PCV2" OR "porcine cytomegalovirus" OR "PCMV"OR "Aujeszky's disease virus" OR "ADV"OR "Encephalmyocarditis Virus" OR "hemagglutinating encephalomyelitis virus" The asterisk is a wildcard search operator to include different ending of the word like "farming" and "farms".

Results
The total number of papers identified was 1469, of which 631 were duplicates, leaving 838 studies.After screening the titles and abstracts, 148 papers were selected for full-text screening.Among these, 49 papers covering seven diseases were deemed eligible for inclusion in the review (Figure 1).R 0 values were estimated for disease transmission within herds/farms and between herds/farms (Figure 2).Various methods were used in these papers to calculate the R 0 values, ranging from complex mathematical approaches like the Next-Generation method to simple direct calculations of the number of secondary infections from simulation models.The most commonly used method to calculate R 0 for a specific disease model was the product of the disease transmission rate and duration of infection (Table 2).These parameters were estimated through model fitting to reported case data or directly inferred from laboratory experiments.
The data used in these studies for calibrating the models primarily consist of surveillance data and laboratory experiments.Surveillance data, as defined in this review, refers to data from a secondary source, such as a government institution in a specific geographic region.In contrast, laboratory experiments were conducted to investigate disease transmission in a controlled environment.Nine studies did not use empirical data to parameterize their models or did not explicitly specify the sources of the data or values.A phylogenetic study utilized genetic data from previous outbreaks to calculate the reproductive number.The counts and distribution of the data types are shown in Figure 3.             Two commercial farrow-to-finish pig farms (HY1 and HY2) located in two different districts in Hung Yen province, Vietnam were selected immediately after ASF outbreaks were confirmed in the two farms.to data from a secondary source, such as a government institution in a specific geographic region.In contrast, laboratory experiments were conducted to investigate disease transmission in a controlled environment.Nine studies did not use empirical data to parameterize their models or did not explicitly specify the sources of the data or values.A phylogenetic study utilized genetic data from previous outbreaks to calculate the reproductive number.The counts and distribution of the data types are shown in Figure 3.

Aujeszky's Disease
Three publications evaluated the R0 value of Aujeszky's disease (Table 2).Out of these three publications, two developed a generic pig population to model Aujeszky's disease transmission, and one developed a model for Aujeszky's disease outbreaks in the Netherlands [27][28][29].All three studies used stochastic compartmental models of disease transmission [27][28][29].The two generic population models estimated the R0 value to be 6 and 4.12, using the Next-Generation Matrix method and direct calculation from model simulations, respectively [27,29].In the Netherlands study, the Next-Generation Matrix method was used to compute the between-herd R0 value, and a survival function method was used for the within-herd R0 value [28].The between-herd R0 value was estimated to range between 1 and 2.5, and the within-herd R0 value was equal to 10 [28].The withinherd R0 value of 10 was three times as high as the median between-herd R0 value of 3.31 (Table 2).

Porcine Circovirus 3
Only one study met the inclusion criteria for Porcine Circovirus 3. The study by Li et al. identified two major genotypes for PCV3 and estimated the R0 value to be 3.08 (95% credible interval (CrI): 0.82-7.96)and 1.82 (95%CrI: 0.63-4.82)for PCV3a and PCV3b, respectively [10].The phylogenetic analysis used for R0 value calculations was based on a birth-death skyline model, and R0 was calculated as the product of the transmission rate and duration of infection [10].

Aujeszky's Disease
Three publications evaluated the R 0 value of Aujeszky's disease (Table 2).Out of these three publications, two developed a generic pig population to model Aujeszky's disease transmission, and one developed a model for Aujeszky's disease outbreaks in the Netherlands [27][28][29].All three studies used stochastic compartmental models of disease transmission [27][28][29].The two generic population models estimated the R 0 value to be 6 and 4.12, using the Next-Generation Matrix method and direct calculation from model simulations, respectively [27,29].In the Netherlands study, the Next-Generation Matrix method was used to compute the between-herd R 0 value, and a survival function method was used for the within-herd R 0 value [28].The between-herd R 0 value was estimated to range between 1 and 2.5, and the within-herd R 0 value was equal to 10 [28].The withinherd R 0 value of 10 was three times as high as the median between-herd R 0 value of 3.31 (Table 2).

Porcine Circovirus 3
Only one study met the inclusion criteria for Porcine Circovirus 3. The study by Li et al. identified two major genotypes for PCV3 and estimated the R 0 value to be 3.08 (95% credible interval (CrI): 0.82-7.96)and 1.82 (95% CrI: 0.63-4.82)for PCV3a and PCV3b, respectively [10].The phylogenetic analysis used for R 0 value calculations was based on a birth-death skyline model, and R 0 was calculated as the product of the transmission rate and duration of infection [10].

Porcine Circovirus 2
Three studies from the same research team estimated R 0 for Porcine Circovirus 2 (PCV2).Andraud et al. calculated the R 0 value within and between pens for both types of within-herd transmission [30].They calculated R 0 under two different assumptions about infectiousness: (1) Infectiousness ends at seroconversion, and (2) Infectiousness ends at seroconversion, and viral load decline occurs.Under assumption 1, the betweenpen R 0 was estimated to be 0.58 (95% Confidence Interval (CI): 0.23-1.47),and under assumption 2, between-pen R 0 increased to 1.2 (95% CI: 0.5, 2.9).The within-pen R 0 was estimated to be 5.5 (95% CI: 3.3-9.0)under assumption 1 and 8.9 (95% CI: 5.1-15.4)under assumption 2 [30].The study developed a stochastic compartmental model for PCV2 transmission parameterized using animal experiment data.R 0 was calculated as the product of the transmission rate and duration of infection [30].Another study by Andraud et al. used a deterministic model, parameterized using animal experiment data, to Vet.Sci.2024, 11, 300 16 of 25 estimate the within-herd transmission of PCV2 [31].The R 0 value was 5.9 (95% CI: 1.8-10.1),estimated using a survival function method [31].The last study used two experimental datasets to estimate the within-herd R 0 value using a final-size algorithm [32].These were time-series individual-level virological data collected from laboratory experiments to investigate PCV2 transmission in controlled environments.The first study was conducted by Rose et al., and the second by Andraud et al. [30,32].Using the data produced by their experiment, Rose et al. calculated R 0 as 6.94 (95% CI: 0.42-15.01)[32].By merging their data with Andruad et al., a previous experiment by the same research group, R 0 was estimated as 5.1 (95% CI: 2.5-8.2) [30,32].

Influenza A
Three publications estimated the within-herd R 0 value of swine Influenza A (Table 2).Rose et al. used case data from outbreaks on swine farms in Brittany, France, and deterministic compartmental models to investigate the dynamics of within-farm swine Influenza outbreaks [5].They used Poisson regression to fit their models to case incidence data and an exponential growth method to calculate the R 0 value.The within-herd R 0 value was estimated to range from 2.5 to 6.9 [5].The remaining two papers used similar study designs and generated comparable results.These similarities include using data from animal experimental trial studies, stochastic compartmental models for disease transmission, and calculating R 0 as the product of the transmission rate times the duration of infection (Table 2).Their R 0 values of 10.4 and 10.66, respectively [33,34], were significantly lower than those reported by Rose et al. [5].

Porcine Reproductive and Respiratory Syndrome Virus
For Porcine Reproductive and Respiratory Syndrome Virus (PRRSV), six publications estimated the within-herd R 0 to range from 2.6 to 5.42 (Table 2).Four studies used a generic pig population to model PRRSV outbreaks, and two modeled PRRSV outbreaks in the United States (Table 2).Of the generic population studies, three used laboratory experiments to inform parameter values of compartmental models and calculated R 0 as the product of the transmission rate times the duration of the infection.The fourth study used the survival function method to calculate within-herd R 0 .
Two studies estimated the between-herd R 0 in the United States using data from the Morrison Swine Health Monitoring Project (MSHMP) and transmission trees to model disease spread [35,36].Pamornchianavakul et al. used phylogenetic analysis, while Arruda et al. utilized the Wallinga and Teunis method to calculate R 0 values [35,36].Both estimated between-herd R 0 values of around 1 despite using different calculation methods (Table 2 [35,36]).

Encephalomyocarditis
Two publications estimated the within-herd R 0 for encephalomyocarditis.Both studies used a stochastic compartmental model to investigate within-farm disease transmission and the White and Pagano method to compute R 0 (Table 2).The first study by Kluivers et al. used serological data from infected pens on a pig farm during an autumn 2001 outbreak in Belgium to inform their model [13,14], and the second study by Maurice et

Classic Swine Fever
Twelve publications evaluated the R 0 value of Classical Swine Fever (CSF) (Table 2).Eight of the twelve papers used generic populations to model CSF transmission; two papers used case data from the Netherlands, one used data from India, and one used data from Bangladesh (Table 2).Six studies calculated R 0 as the product of transmission rate times duration of infection; two used an attack rate method, two used an exponential growth method, and two used a martingale estimator method (Table 2).
A machine learning model was developed by Suresh et al. to create an early warning system for potential outbreaks related to climate change [42].The model was trained on geotagged incidence data for Guwahati City, Assam, from 2005 to 2021 [42].Risk factors, such as humidity and vegetation, were added to the model to ascertain the climatedisease relationship on a spatial grid.R 0 was calculated using the attack rate for each grid, producing values ranging from 1.04 to 2.07 [42].Chowdhuary et al. used an exponential growth function to model surveillance data for a 2015 outbreak in the Kurigram district of Bangladesh [41].They used two methods for calculating R 0 : the attack rate and the exponential growth rate (Table 2).The attack rate method estimated R 0 as 1.6 (95% CI: 1.5-1.7)and exponential growth estimated as 1.5 (95% CI: 1.3-1.7)(Table 2).
Stegeman et al. published two papers using data from the 1997 to 1998 outbreak of CSF in the Netherlands to estimate the within-and between-herd R 0 (Table 2).The between-herd R 0 was estimated to be 6.8 [43].Data were obtained from 21,500 herds in the Netherlands, of which 429 experienced an outbreak.In the second paper, a simple deterministic compartmental model was fitted to the data, and the within-herd R 0 was calculated as the product of the transmission rate times the duration of infection.To calculate the within-herd R 0 , Stegeman et al. used data from 82 of the 429 infected herds because these outbreaks could be identified from a single known contact [44].The withinherd R 0 was estimated at 2.9, using the exponential growth method applied to a stochastic compartmental model [44].
Durand et al. investigated the potential use of clinical data (quantitative and qualitative), considered as infectivity markers, to reliably estimate CSF R 0 value.The clinicalqualitative dataset reported the number of infected animals with clinical scores greater than 0, and the clinical-quantitative dataset reported the sum of clinical scores of infected animals [48].They used viremia-based data (virus in the bloodstream) as the standard dataset for R 0 calculation [48].R 0 values computed from the clinical datasets were higher than the viremia-based ones, but the confidence intervals all overlapped, indicating that these values were not significantly different (Table 2).The within-herd values varied from 4.0 (95% CI: 1.9-8.4) to 12.2 (95% CI: 5.5-27.3),and the between-herd values varied from 1.1 (95% CI: 0.4-2.9) to 1.8 (95% CI: 0.7-4.8)(Table 2) [48].
Laevans et al. performed two animal experimental studies using pigs of different ages to evaluate the within-pen risk of CSF transmission [25,26].They used experimental data with different final outbreak size methods for R 0 calculation to estimate within-herd R 0 values [18,19].The weaner pigs had an estimated R 0 of 81.3, whereas the slaughter pigs had a significantly smaller estimate of 13.7 [18,19].Klinkenberg et al. used data collected in the Laevans et al. experiments to parameterize a stochastic compartmental model of CSF transmission [18,19,45].Calculating R 0 as the product of the transmission rate times duration of infection, they estimated the within-herd R 0 to be 100 for weaner pigs and 15.5 for the slaughter pigs; these estimates were much higher than those of Laevans et al. (Table 2).These results clearly illustrate the high sensitivity of R 0 to the calculation method.
Weesendrop et al. reported the highest within-herd R 0 values of 143 for a high transmissibility strain and 23 for a low transmissibility strain [49].Bouma et al. reported the lowest within-herd R 0 value, 2.3 (95% CI: 0-5.0) [46].This disease had the largest range of R 0 estimates across all diseases, likely due to the diversity in the R 0 calculation methods and the type of data used.Studies using laboratory experimental data were more likely to generate higher R 0 values than those using surveillance data (Table 2).

African Swine Fever
Eighteen publications estimated R 0 for African Swine Fever (ASF), the most common disease identified in this search (Table 2).Among the articles extracted, there was an even split between studies using generic populations to model ASF transmission and those using laboratory experiment data and location-specific surveillance data.Nine studies utilized surveillance data to calibrate their models, four studies used data from Vietnam, two studies used data from China, two studies used data from Russia, one study used data from Ukraine, and one study used data from Uganda (Table 2).Fourteen studies estimated within-herd R 0 , and five studies estimated between-herd R 0 .Using genotyping information from surveillance data, R 0 results were stratified by genotype.

Genotype IX
Only Barongo et al. investigated genotype IX, a strain mostly found in eastern Africa [51].This Uganda study focused on small free-range pig production in the Gulu district, whose economy is rooted in small-scale agriculture [51].The data used in this study were part of a larger effort to understand the transmission of ASF in Gulu District [51].The period of interest was from April 2010 to November 2011 [51].Herds were defined as all pigs in a pig-keeping household [51].Barongo et al. used three separate models to depict transmission: a spatial simulation, an epidemic doubling time method, and a simple SI (Susceptible-Infected) deterministic model.The spatial simulation, referred to as the nearest infectious neighbor by Barongo et al., had the highest estimate of R 0 as 3.24 (95% CI: 3.21-3.27)[51].The doubling time method produced an R 0 of 1.63.For the SI model, R 0 was computed as the product of the transmission rate and duration of infection, and the model was fitted to the data in three ways: linear regression, curve fitting, and bootstrapping, leading to slightly different R 0 values.The within-herd R 0 was estimated as 1.58 for curve fitting, 1.77 for bootstrapping, and 1.9 for linear regression [51].
3.8.2.Genotype I R 0 estimates for genotype I were calculated in two laboratory experiments and in one study using surveillance data from Ukraine.Eble et al. used data from a previous laboratory experiment for the Netherlands/86 strain to calculate the R 0 for carrier animals, finding that transmission does occur but at low levels with R 0 less than 1 (R 0 = 0.3) [52].De Carvalho Ferreira et al. generated R 0 estimates of 18.0 (95% CI: 6.90-46.9)for the Malta/78 isolate and 4.92 (95% CI: 1.45-16.6)for Netherlands/86 [53].The Ukraine study investigated the transmission of ASF in the Odesa region of Ukraine [54].It used historical data from the 1977 outbreak collected by the USSR Ministry of Agriculture [54].The study estimated the between and within-herd R 0 to be 1.65 (95% CI: 1.42-1.88)and 7.46 (95% CI: 5.68-9.21),respectively [54].

Genotype II
The most recent R 0 estimates come from outbreaks in Asia.Of the 11 studies of genotype II ASF, three used data from outbreaks in China and four studies used data from Vietnam.All the studies out of Asia pertain to the introduction of the disease to the country, which occurred from late 2018 to early 2019 [55,[57][58][59][61][62][63].
All studies on ASF in China used deterministic compartmental models and the Next-Generation Matrix method to estimate R 0 .Haung et al. used incidence data from the Ministry of Rural Agriculture of China and estimated R 0 to be 0.6 [55].Song [59].
All four studies pertaining to Vietnam used data from the 2019 outbreak of ASF in the northern region, the introduction of the disease into the country.Three of the four used compartmental models, and one used a serial interval model assuming a Poisson distribution to model daily incidence.Mai et al. was the only study that calculated the between-herd R 0 [57].The between-herd R 0 was calculated for 15, 19, and 30-day infectious periods, yielding a range of values from 1.41 to 10.8 [57].Table 2 shows the between-herd R 0 values at different locations for the 15-day infectious period.Mai et al. used three methods to estimate the within-herd R 0 using data from seven farms in Hung Yen and three farms in Ninh Binh and Ha Nam in Vietnam [62].They estimated R 0 as 1.49 using the exponential growth method, 1.58 using the White and Pagano method, and 1.46 using the attack rate method [62].These data were also used to parameterize a SIR (Susceptible-Infected-Recovered) model under two farm sizes and lengths of the infectious period.They estimated R 0 for a 100-299 pig farm to be 1.66 (95% CI: 0.88-2.84)and 1.4 (95% CI: 1.01-1.90)for a 300-999 pig farm [62].Another study used only two mid-sized farms from Hung Yen using a statistics-based model [57].These facilities were chosen for commercial design, strict biosecurity measures, and the way workers were assigned to only work in one barn with no visitors allowed on the premises [63].Le et al. collected data from 15 March through 15 November 2019, on a small family-owned farm in Hanoi, Vietnam, that only had seventeen pigs [61].They estimated the within-herd R 0 to be 10.4 (95% CrI: 1.1-30.4)[61].The remaining two studies estimated R 0 to range from 1.55-1.78[57,62].
The two studies from Russia used compartmental models and data from the Caucasian region and calculated R 0 as the product of the transmission rate and the duration of the infection.Guinat et al. used data from the Federal Research Center for Virology and Microbiology and selected data from nine industrial herds between 2010 and 2014 [60].A stochastic modeling approach was implemented to estimate the within-herd transmission of the infection [60].The smallest within-herd R 0 median value was estimated as 4.4 (95% CrI: 2.09-13.4)and the largest as 17.3 (95% 3.5-45.5)[60].Gulenkin et al. obtained data from the World Animal Health Information System Office International Epizootic (WAHID OIE) database for the period between 2007 and 2010 [56].They used a deterministic model for both between and within-herd disease transmission.Using a curve fitting method to estimate the model parameters, the within-herd R 0 was estimated to range from 8 to 11.Alternatively, using a general linear model for model fitting R 0 , the value was estimated as 9.8 (95% CI: 3.9-15.6)[56].Estimates for the between-herd R 0 were limited to data from the Republic of North Ossetia in 2008, and its value ranged from 2 to 3 [56].
Two studies related to ASF genotype II used laboratory experiments to model withinherd ASF transmission.Guinet et al. utilized compartmental models, and the product of the transmission rate times duration of infection was the most used approach to calculate R 0 .Guinet et al. conducted an experiment on the Georgia 2007/1 strain and used a stochastic model considering different latent periods [64].They calculated R 0 for within and between pens for 3-, 4-, and 5-day latent periods.Oh et al. inoculated pigs using ASF from a 2020 outbreak in Thanh Hóa province, Vietnam, using survival analysis to model transmission between pigs [65].R 0 was calculated using the exponential growth method and White and Pagano method, producing the values 2.91 (95% CI: 1.51-7.06)and 4.015 (95% CI; 1.13-9.8),respectively.

Genotype Not Specified
Five papers produced mathematical models developed to analyze the dynamics of ASF outbreaks but did not specify the genotype used in parameterization.Parameters for these models come from the literature; therefore, these models are proof of concept rather than applied to specific outbreaks [68][69][70].Four of the five studies estimated within-herd R 0 ranging from 3.77 to 13.02.The fifth study developed a model to analyze the spread of ASF caused by a contaminated human vector between farms, calculating R 0 as 18.57 [66].

R 0 Methodologies
The methods used for calculating R 0 varied greatly between studies.This review found 16 different ways R 0 was calculated, with each disease having at least two different calculation methods.Calculation methods ranged from simple, like the product of the transmission rate and the duration of infection, to more complex methods, like the Next-Generation Matrix.The measure of R 0 itself is situation-specific.Studies using data from animal experimental trials tended to use the product of the transmission rate times the duration of the infection, while studies using surveillance data employed the largest range of calculation methods, with most studies using more than one method to estimate R 0 .For example, Chowdhuary et al. used both attack rate and exponential growth methods to estimate the R 0 for CSF and found no significant differences between the estimates [41].On the other hand, Barongo et al. used four methods for calculating R 0 , with one of the estimates above 3 and the rest between 1.5 and 2 [51].Modeling studies using complex compartmental models for disease transmission have tended to use the Next-Generation Matrix method for R 0 calculation.Differences in R 0 estimates for the same pathogen may vary with R 0 calculation methods, pathogen variants [37,71], the data source used for model calibration (surveillance vs. laboratory data), and data location.

Comparison of R 0 Estimates from Laboratory and Surveillance Data
Laboratory experiments often yield higher R 0 estimates compared to surveillance data for several reasons.It must be noted that laboratory experiments are limited in scope, as they can primarily be used to calculate within-herd transmission for a relatively small number of animals.In contrast, surveillance data are broad and can encompass not only the incidence on one farm but an entire geographic area, covering hundreds or thousands of animals.Due to the magnitude of farms, surveillance data may underreport cases as animals are not continuously tested, whereas the controlled environment of laboratory experiments allows for animals to be continuously tested.Furthermore, the environment of laboratory experiments lends itself to an increased transmission risk.Experiments are often conducted in relatively small pens or rooms, potentially enhancing disease transmission compared to actual farms that generally have larger pens or rooms.Laevans et al. attributed differences in pen size for weaner and slaughter pigs as the primary factor leading to a significant magnitude difference between the two estimates, 81.3 for weaner pigs and 13.7 for slaughter pigs [18,19].Additionally, in experimental trials, the control group is not subject to any disease mitigation measures, whereas farms, especially industrial ones, continuously employ some level of biosecurity or mitigation measures.Some experiments in this review utilized high-virulence strains, which may not reflect the strains currently circulating on farms, potentially leading to an overestimation of R 0 and misrepresentation of actual transmission risks [49,50,53].

Comparison of within and Between-Herd R 0
Understanding the transmission dynamics of viruses involves examining two main aspects: within-and between-herd R 0 .Between-herd R 0 tends to show lower values compared to within-herd R 0 , partly due to how it is measured.The between-herd R 0 estimates the virus's ability to spread from one farm to another, essentially tracking its movement between different populations in distinct areas.The unit of measurement here is a farm, regardless of its pig population size.Conversely, within-herd R 0 focuses on the virus's transmission from pig to pig within a single farm.This often yields higher estimates, possibly because there are more opportunities for interactions among animals within the same farm compared to interactions between animals from different farms, such as through animal movement.Consequently, diseases like viral respiratory illnesses are more likely to spread rapidly within a single farm than between different farms.When calculating within-herd R 0 , we typically rely on data about the incidence of infection within a specific farm, while between-herd R 0 is based on information about the number of farms infected over time.

Disease Eradication and R 0
The basic reproduction number is a key epidemiological metric that can be used to measure the transmission risk of infectious disease, predict the expected size of an outbreak, and estimate the control effort needed to prevent or interrupt disease transmission, among others [72].However, the R 0 value is not necessarily a direct indication of the potential eradication of an animal infectious disease.Disease eradication depends on several factors, such as disease characteristics (e.g., host reservoirs, clinical symptoms, infectious period, and immunity duration), economic factors, political will, and control tools (e.g., vaccine, depopulation, culling, movement restrictions) [73,74].For example, an animal disease with a low R 0 value, several host species, and/or wildlife reservoirs, such as Circovirus infection, may be harder to eradicate than a disease with a higher R 0 value but limited to a single host species with no persistence in wildlife, such as AD.

Limitations of R 0 Calculation for Complex Diseases
Diseases such as PRRSV have a complex infectious period, during which infectious hosts experience both acute and subclinical/chronic infectious phases with different levels of infectivity.Although this characteristic is central to PRRSV disease progression and transmission risk, most PRRSV R 0 models have not explicitly considered the differential role of these infectious phases in disease transmission, and none has estimated their contribution to R 0 value [7,[35][36][37][38][39][40]75,76].Such information would be pivotal for accurately evaluating the timing and effectiveness of disease control strategies.Further studies should investigate the differential contribution of acute and subclinical/chronic infectious phases to R 0 and develop a general framework to deal with these situations in computing R 0 and other disease transmission risk metrics.

Conclusions
Despite the limitations of R 0 , it is a useful tool to begin quantifying the risk of disease outbreaks and the magnitude of an outbreak and investigate intervention strategies' effectiveness for disease control or prevention.For example, understanding the R 0 for within-herd transmission is crucial for determining the vaccination efficacy required for preventing or eradicating disease on a farm.Between-herd R 0 is paramount for understanding the vulnerability of farm networks to disease outbreaks and the vaccination coverage required to prevent large-scale disease outbreaks.Between-herd transmission is mostly driven by interactions between farms through commerce, supply chains, and the movement of infected livestock and related products.To improve the estimate of between-herd R 0 , future studies should explicitly incorporate data on interactions between farms.However, such interaction and mobility datasets are not readily available for most countries.
Swine diseases with a larger R 0 emphasize the urgency of promptly detecting and responding to outbreaks.Addressing outbreaks can be challenging, regardless of resource availability.Through timely case surveillance, forecasting tools, and rapid dissemination of information to farmers, veterinarians, and government officials, animals and the food supply can be better protected against infectious disease outbreaks.Preventing swine respiratory diseases from spreading and entering the global supply chain is imperative, as they pose an imminent threat to food supply.Continuous monitoring and evaluation of pathogen transmissibility are paramount to enhance disease surveillance efficiency and decrease the impact of livestock diseases.

Figure 2 .
Figure 2. Distribution of R0 values by disease for within-and between-herd transmissio

Figure 2 .
Figure 2. Distribution of R0 values by disease for within-and between-herd transmission.Figure 2. Distribution of R 0 values by disease for within-and between-herd transmission.

Figure 2 .
Figure 2. Distribution of R0 values by disease for within-and between-herd transmission.Figure 2. Distribution of R 0 values by disease for within-and between-herd transmission.

Figure 3 .
Figure 3. Distribution of data types extracted for all swine respiratory diseases.

Figure 3 .
Figure 3. Distribution of data types extracted for all swine respiratory diseases.

Table 1 .
Terms by category used for generating results.

Table 2 .
Extraction results stratified by disease.