Influenza A(H1N1)pdm09 virus infection in Norwegian pigs was largely subclinical. This study tested the hypothesis that the infection causes negligible impact on pigs’ growth performance in terms of feed conversion efficiency, daily feed intake, daily growth, age on reaching 100 kg bodyweight and overall feed intake. A sample of 1955 pigs originating from 43 breeding herds was classified into five infection status groups; seronegative pigs (n = 887); seropositive pigs (n = 874); pigs positive for virus at bodyweight between 33 kg and 60 kg (n = 123); pigs positive for virus at bodyweight between 61 kg and 80 kg (n = 34) and pigs positive for virus at bodyweight between 81 kg and 100 kg (n = 37). Each pig had daily recordings of feed intake and bodyweight from 33 kg to 100 kg. Marginal effects of the virus infection on the outcomes were estimated by multi-level linear regression, which accounted for known fixed effects (breed, birthdate, average daily feed intake and growth phase) and random effects (cluster effects of pig and herd). The seropositive and virus positive pigs had decreased (P value<0.05) growth performance compared to seronegative pigs even though feed intake was not decreased. Reduced feed conversion efficiency led to lower average daily growth, additional feed requirement and longer time needed to reach the 100 kg bodyweight. The effects were more marked (P value<0.03) in pigs infected at a younger age and lasted a longer period. Despite increased feed intake observed, their growth rates were lower and they took more time to reach 100 kg bodyweight compared to the seronegative pigs. Our study rejected the null hypothesis that the virus infection had negligible adverse effects on growth performance of Norwegian pigs.

Newly emerged influenza A(H1N1)pdm09 virus infection in Norwegian pigs, although often observed in a subclinical form, can lower the pig's growth performance by reducing feed efficiency in terms of a poorer feed conversion ratio. Infected pigs would consume more feed and require protracted production time to reach market weight. In our observational longitudinal study, growth performance data from 728 control pigs and 193 infected pigs with known viral shedding time points were analyzed using mixed linear regression models to give estimates of the marginal effects of infection. Gaussian curves describing the variability of the estimates at the individual pig level formed the fundamental inputs to our stochastic models. The models were constructed to simulate the summed negative effects of the infection at the batch level of 150 fattening pigs growing from 33 to 100 kg. Other inputs of variability and uncertainty were 1) batch transmission points, 2) pig infection points to reflect the disease transmission dynamics of the virus, and 3) final prevalence of infected pigs in the batch. Monte Carlo random sampling gave 5,000 estimates on the outputs of the marginal effects for each pig. These results were summed up to provide estimates for a batch size of 150 pigs. This figure was adjusted by our final prevalence distribution function, which was also derived from the longitudinal study with 12 cohorts of infected pigs. For a 150-fattening-pig herd randomly selected from the population, the marginal effects of the infection were 1) 835 kg (fifth percentile) to 1,350 kg (95th percentile) increased feed intake and 2) 194 (fifth percentile) to 334 (95th percentile) pig days in excess of expected figures for an uninfected batch. A batch infected during growth phase 3 (81 to 100 kg BW) gave the worst results since the longitudinal study showed that a pig infected during growth phase 3 required more feed and a greater protracted production time compared to younger infected pigs. Sensitivity analysis showed that final prevalence had the greatest impact on the conditional mean and variation of the marginal effects of infections. Batch transmission point was the next most influential factor. Lowering the final prevalence and preventing older fattening pigs from being infected will have the greatest benefit in saving feed cost and reducing delay in getting the pigs to the market.

SUMMARY: The incursion of influenza A(H1N1)pdm09 virus was detected by Norway's active serosurveillance of its pig population in 2009. Since then, surveillance data from 2010 to 2014 revealed that 54% of 5643 herd tests involving 1567 pig herds and 28% of 23 036 blood samples screened positive for antibodies against influenza A virus. Positive herds were confirmed to have influenza A(H1N1)pdm09 virus infection by haemagglutination inhibition test. In 50% of positive herd tests, ⩾60% of the sampled pigs in each herd had antibodies against influenza A(H1N1)pdm09 virus. This within-herd animal seroprevalence did not vary for type of production, herd size or year of test. The overall running mean of national herd seroprevalence, and annual herd incidence risks fluctuated narrowly around the means of 45% and 32%, respectively, with the highest levels recorded in the three densest pig-producing counties. The probability of a herd being seropositive varied in the five production classes, which were sow pools, multiplier herds, conventional sow herds, nucleus herds, and fattening herds in descending order of likelihood. Large herds were more likely to be seropositive. Seropositive herds were highly likely to be seropositive the following year. The study shows that influenza A(H1N1)pdm09 virus is established in the Norwegian pig population with recurrent and new herd infections every year with the national herd seroprevalence in 2014 hovering at around 43% (95% confidence interval 40–46%).

Since the incursion of influenza A(H1N1)pdm09 virus in 2009, serosurveillance every year of the Norwegian pig population revealed the herd prevalence for influenza A(H1N1)pdm09 (HIN1pdm09) has stabilised between 40% and 50%. Between 30 September 2009 and 14 September 2017, the Norwegian Veterinary Institute and Norwegian Food Safety Authority screened 35,551 pigs for antibodies to influenza A viruses (IAVs) from 8,636 herds and found 26% or 8,819 pigs' sera ELISA positive (titre ≥40). Subtyping these IAV antibodies from 8,214 pigs in 3,629 herds, by a routine haemagglutination inhibition test (HAIT) against four standard antigens produced 13,771 positive results (HAIT titre ≥40) of binding antibodies. The four antigen subtypes eliciting positive HAIT titre in descending frequencies were immunogen H1N1pdm09 (n = 8,200 or 99.8%), swine influenza A virus (SIVs) subtypes swH1N1 (n = 5,164 or 62%), swH1N2 (n = 395 or 5%) and swH3N2 (n = 12 or 0.1%). Of these 8,214 pig pigs sera, 3,039 produced homologous HAIT subtyping, almost exclusively immunogen H1N1pdm09 (n = 3,026 or 99.6%). Using HAIT titre of pig and herd geometric mean titre (GMT) as two continuous outcome variables, and with the data already structured hierarchically, we used mixed effects linear regression analysis to investigate the impact of predictors of interests had on the outcomes. For the full data, the predictors in the regression model include categorical predictors antigen subtype (H1N1pdm09, swH1N1, swH1N2 & swH3N2), and production type (sow herd or fattening herd), ordinal predictors year (longitudinally from 2009 to 2017) and number of antigens in heterologous reactions (1, 2, 3, 4) in the same pig serum. The last predictor, the proportion of HAIT positive (antigen specific) in tested pigs within the herd, was a continuous predictor, which served as a proxy for days post-infection (dpi) or humoral response time in the pig or herd. Regression analysis on individual pig HAIT titres showed that antigen as a predictor, the coefficient for immunogen H1N1pdm09 was at least fourfold higher (P < 0.001) than the three SIVs antigen subtypes, whose much lower coefficients were statistically no different between the three SIVs antigen subtypes. Correspondingly, for herd GMT, immunogen H1N1pdm09 was 28–40-fold higher than the three SIVs antigen subtypes. Excluding the HAIT data of the three SIVs antigen subtypes, regression analysis focusing only on immunogen H1N1pdm09 increased greatly the coefficients of the predictors in the models. Homologous reactions (99.6% H1N1pdm09) have lower HAIT titres while the likelihood of the number of antigens involved in HAIT heterologous reactions in a single pig serum increased with higher HAIT titres of immunogen H1N1pdm09. For predictor ‘production’, sows and sow herds had higher HAIT titres and GMT compared to fattening pigs and fattening herds respectively. Herds with ‘higher proportion of pigs tested positive’ also had higher HAIT titre in the pig and herd GMT.

In order to quantify projections of disease burden and to prioritise disease control strategies in the animal population, good mathematical modelling of infectious disease dynamics is required. This article investigates the suitability of discrete-time Markov chain (DTMC) as one such model for forecasting disease burden in the Norwegian pig population after the incursion of influenza A(H1N1)pdm09 virus (H1N1pdm09) in Norwegian pigs in 2009. By the year-end, Norway’s active surveillance further detected 20 positive herds from 54 random pig herds, giving an estimated initial population prevalence of 37% (95% CI 25–52). Since then, Norway’s yearly surveillance of pig herd prevalence has given this study 11 years of data from 2009 to 2020 to work with. Longitudinally, the pig herd prevalence for H1N1pdm09 rose sharply to >40% in three years and then fluctuated narrowly between 48% and 49% for 6 years before declining. This initial longitudinal pattern in herd prevalence from 2009 to 2016 inspired this study to test the steady-state discrete-time Markov chain model in forecasting disease prevalence. With the pig herd as the unit of analysis, the parameters for DTMC came from the initial two years of surveillance data after the outbreak, namely vector prevalence, first herd incidence and recovery rates. The latter two probabilities formed the fixed probability transition matrix for use in a discrete-time Markov chain (DTMC) that is quite similar to another compartmental model, the susceptible–infected–susceptible (SIS) model. These DTMC of predicted prevalence (DTMCP) showed good congruence (Pearson correlation = 0.88) with the subsequently observed herd prevalence for seven years from 2010 to 2016. While the DTMCP converged to the stationary (endemic) state of 48% in 2012, after three time steps, the observed prevalence declined instead from 48% after 2016 to 25% in 2018 before rising to 29% in 2020. A sudden plunge in H1N1pdm09 prevalence amongst Norwegians during the 2016/2017 human flu season may have had a knock-on effect in reducing the force of infection in pig herds in Norway. This paper endeavours to present the discrete-time Markov chain (DTMC) as a feasible but limited tool in forecasting the sequence of a predicted infectious disease’s prevalence after it’s incursion as an exotic disease.