3.2 Descriptive characteristics of the 29 studies
Descriptive information about the studies is included in Table 2. Results indicated that 69% (20/29) of the literature on PoAstVs were published between 2011 and 2018. A higher number of studies (n = 13) were conducted in Europe compared to Asia (n = 7) or North America (n = 9). Most studies were conducted after 2010. We did not find any relevant paper having a group of pigs (i.e., population) as a unit of analysis. Almost all of the relevant papers (n = 29) explained data at the level of the individual on exposure and disease, including cross-sectional study, case-control study, and case report (Table 3).
Thirteen of the 14 case-control studies were based on the disease status of pigs and was matched by the age of the pigs except in one study from Cai et al., 2016. In 12 case-control studies, cases were defined as pigs with diarrhea; in 2 studies, cases were pigs with neurological signs. The controls were clinically healthy pigs. Most of the cases and controls were subjective except in Boros et al., 2017 and Blomström et al., 2014 where cases were selected objectively based on the severity of the disease. Three of the case-control studies used herd (Goecke et al., 2017), region (Xiao et al., 2013), and country (Zhou et al., 2016) as strata to balance the data set. There was no experimental or challenge trial study, or studies describing the pathogenesis of PoAstVs. Overall detection was calculated for each of the cross-sectional studies (n = 13) and case reports (n = 2), and was divided by age groups, sample types and diagnostic tool used (Table 4).
The studies included multiple age groups representing different stages of swine production including suckling piglets (< 28 days), nursery pigs (4-9 weeks), finisher or market hogs (9-25 weeks), gilts, sows, boars, and unknown ages. The frequency of detection of PoAstVs ranged from 0 to 100% in pigs of all ages from suckling to adults. The overall detection across studies (Table 4) by age group was 22% (230 of 1045) in suckling piglets, 67% (131 of 197) in nursery, 59% (154 of 263) in finisher, 36% (13 of 36) in gilts, 37% (33 of 89) in sows, 82% (9 of 11) in boar and 65% (573 of 881) in unknown age indicating a higher detection proportion in growing pig in comparison with breeding sows. These suggested growing pigs were the best age group to surveil PoAstVs in the pig population. Boars had an 82% detection rate, but with limited (n = 11) sample size, which may be due to selection bias.
Nine different sample types were identified for the detection of PoAstVs including fecal, CNS, serum, nasal, oral fluids, feeders, environmental (interior), environmental (exterior), environmental (abattoir) and livestock transport vehicle samples. The overall detection rate across studies in Table 4 by sample types was 38% (770 of 2016) in fecal samples, 80% (8 of 10) CNS samples, 4% (7 of 180) serum samples, 21% (20 of 95) nasal samples, 86% (73 of 85) in environmental samples (interior), 24% (15 of 63) in environmental samples (exterior), 69% (68 of 99) in environmental samples (abattoir), 42% (34 of 82) in livestock transport vehicle samples and 100% in (12 of 12) in feeders. This suggests feeders, interior environmental samples, and CNS samples are the sample types with higher sensitivity in comparison with other sample types. Although the sample size for feeders and CNS samples were limited which may be due to selection bias. The serum samples had a poor detection rate of 4% (7 of 180), indicative of transient viremia in pigs infected with PoAstVs. Only one study used nasal samples for detection of PoAstV specifically PoAstV4 and investigated its association with respiratory disease. The detection with nasal samples was found to be 21% (20 of 95) with significantly lower Ct value than fecal samples at P = 0.04. The Ct value ranged from 19.5 to 35.4 in positive nasal samples compared to range of 24.3 to 36.3 in fecal samples positive by RT-PCR.
Out of 29 relevant papers, 48% (14 of 29) were case-control studies. The odds ratio (OR) ranged from 0.25 to 36 and 95% confidence interval (CI) was calculated for each of the studies and was sorted from lower to higher OR (Table 5). Seventy-two percent (21 of 29) of the studies reported the detection of other agents with viral etiology. Moreover, 83% (24 of 29) of the studies did not report conducting bacterial culture, 80% (4 out of 5) of the studies where bacterial culture was done were negative although one of the studies found Escherichia coli described as from non-relevant serotype. The differential diagnosis of PoAstVs with different viral agents was shown in Table 6. PCR-based assays were the most common diagnostic tool used for the diagnosis of PoAstVs. 81% (17 of 21) studies used PCR to differentiate with other viral agents and the remaining 19% (4 of 21) used different types of sequencing tools including phylogenetic analysis, Illumina analysis, and metagenomic sequencing. Only 10% (3 of 29) of the studies assessed histologic lesions. Two of those three reported histologic lesions including mononuclear perivascular cuffing with vasculitis, neurophagia, multifocal microgliosis, and severe nonsuppurative polioencephalomyelitis and associated detection of PoAstV more specifically PoAstV3 with neurologic disease (Arruda et al., 2017; Boros et al., 2017). The remaining study reported histologic lesions including mild to moderate vacuolar changes of white matter and associated detection of PoAstVs with congenital tremor (Blomström et al., 2014). None of the case-control studies describing cases as pigs with diarrhea (11 of 14) showed a link between detection of PoAstVs with histologic examination of gastroenterologic tissues including gastritis, atrophic enteritis or colitis. These findings demonstrate the need for improvement in the diagnosis of PoAstVs-associated diseases (Shan et al., 2011; Xiao et al., 2013; Cai et al., 2016; Goecke et al., 2017; Zhou et al., 2016; Ito et al., 2017; Salamunova et al., 2018; Lee et al., 2013; Monini et al., 2015; Kumthip et al., 2018; Xiao et al., 2017; Wallgren et al., 2014). One of the twenty-eight studies reported gross lesions (Schiavon et al., 2016) including catarrhal enteritis with foam, loose tone of the intestinal tract, peritonitis, edema of colon and hyperemia of pyloric region of the gut.
Here we summarized the literature, identified by the scoping review to determine whether each of the nine Bradford Hills criteria was fulfilled.