Animal health surveillance activities in Cambodia are limited with only a small number of neglected parasitic diseases and FMD surveillance programs supported by the government. Other programs including influenza (Goutard et al., 2015; Osbjer et al., 2017), Japanese Encephalitis, West Nile (Auerswald et al., 2020) and wildlife disease (Hul et al., 2021) surveys have been supported by international organizations and/or aid agencies. To support the animal health service to strengthen disease surveillance activities, our study chose abattoir-based surveillance over structured surveillance as it is simpler to implement in a low-resource setting. Even though disease information collected from an abattoir surveillance program has limitations and may not represent the true prevalence of diseases in the population (Cannon & Roe, 1982), it is still useful to provide an indication of the likely prevalence and therefore probable impact of diseases, as well as indications of geographic distribution, for further investigation. Important in this context, a study by Blacksell et al. (2008) also demonstrated that overall FMD seropositivity detected by structured and abattoir surveys were relatively similar.
A major constraint identified during our study was that central staff from NAHPRI were well-trained and often facilitated field sample collections with little assistances from abattoir veterinarians, resulting in limited knowledge transfer and capacity building of on-site staff. Another constraint was the lack of laboratory capability and capacity due to inadequate human and financial resources. Capacity building of field provincial officers (i.e. abattoir veterinarians, animal health workers, etc.) in disease surveillance, sample collection, submission, case reporting and biosafety principles are critical for early detection and disease monitoring and control. Currently, many field investigations are performed by NAHPRI staff. The average cost of the field consumables per sample in Cambodia was half of the cost of those previously reported in a similar program in Lao PDR (Siengsanan-Lamont et al., 2021). Local supplies are widely available in Cambodia resulting in the lower cost of consumables. However, in looking at cost-effectiveness and sustainability the cost of the diagnostic test kits must also be considered. In this instance, specific project resources supported the surveillance in order to obtain baseline indications of prevalence and to guide how to build a system for the longer term. The question arises as to the longer-term utility of such a surveillance system. The system might be used from time to time to get a snapshot of the likely prevalence of priority diseases, especially zoonoses, in the livestock population. In a similar program in Lao PDR, the surveillance activity and associated training has also provided capacity building for field surveillance and laboratory diagnosis and was generally met with positive attitudes of field staff.
The total numbers of samples collected per trip varied depending on the numbers of animals processed for slaughter on the sample collection days. Animals were often delivered to slaughterhouses by traders or delivery drivers who may or may not have had full records of individual animals, but generally did not. Thus, some biodata, especially vaccination history, was not available, but if animals originated from outside the country this was generally known. Interpretation of the results needed to take into account biases caused by these limitations of the sample and data collections. An interesting observation from the study was the significant numbers of animals of Thai origin processed in the Cambodian abattoirs and previous studies have reported similar findings. Cambodia has previously been reported as a thoroughfare for cattle from Myanmar and Thailand to Vietnam and China (Pham et al., 2015). Declining cattle production in Cambodia coupled with an increased demand for animal protein (Olmo et al., 2017) has driven the importation of livestock, with the sample in this study revealing that a large proportion (~ 43%) of cattle processed in these slaughterhouses in Phnom Penh had come in from Thailand. Pisei (2020) reported that Cambodia imported around 20% (~ 2,000–3,000 live pigs/day) of its pigs from Thailand. Another study in 2012 reported that Cambodia imported pigs and cattle from Thailand and cattle from Vietnam (Kerr, Sieng, & Scoizec, 2013). However, our study had no record of animals at the abattoir from Vietnam. The absence of animals from Vietnam was likely due to the first ASF outbreak in Vietnam in February 2019 (Woonwong, Do Tien, & Thanawongnuwech, 2020) at which time the Cambodian government banned the importation of pigs from Vietnam in March 2019 (Xuxin, 2019). There are three large scale commercial piggery companies in Cambodia owned by multinational regional agribusiness companies (Pisei, 2020) which were the sources of most swine samples in our study. In May 2020, the General Directorate of Animal Health and Production at the Ministry of Agriculture, Forestry and Fisheries announced a reduction of live pigs imports from neighbouring countries, mainly from Thailand, to 1,800-2,100 head per day and a prohibition of the transit of live pigs from Thailand to Vietnam, to help support local pig production (Chan, 2020). As demand for red meat continues to grow (Woonwong, Do Tien, & Thanawongnuwech, 2020; Young et al., 2014), movements of animals and animal products continue to pose a risk for spreading transboundary animal diseases. It is a major challenge for Cambodian authorities to maintain meat supplies and at the same time prevent the movement of serious livestock diseases into the country. Currently, ASF poses a major threat to pig production in Cambodia, especially at the smallholder level in villages. And lumpy skin disease (LSD) is also threatening the cattle population in the region as it has become widespread in China (Roche et al., 2021), and recent outbreaks have been reported in Thailand (Sripiachai, 2021). Abattoir surveillance might be useful to monitor the prevalence of endemic disease conditions, or to quickly establish the distribution and impact of a recently introduced disease (e.g. PRRS). And it may help detect a new disease with less dramatic clinical manifestations, such as LSD.
In this study, the seroprevalence of Q fever and brucellosis in cattle was relatively low. In 2008, 120 cattle samples collected from six villages in three provinces in Cambodia tested negative to brucellosis by RBT (Sothoeun, Young, & Windsor, 2013). Another study conducted in 2015 in Sa Kaeo province, Thailand (located close to the Thai-Cambodian border) reported that the herd-level seroprevalence of brucellosis and Q fever of beef cattle were 2.6% (95% CI 0.9,7.3) and 4.3% (95% CI 1.8, 9.6) (Colombe et al., 2018). Brucellosis and Q fever are zoonoses and pose a public health risk, especially to people who closely contact with infected animals (Mori & Roest, 2018). Further investigation in Takeo province where seropositive animals originated could provide more information on the disease distribution. It is interesting that both these diseases have a very low seroprevalence in Cambodia, where there are no control programs in place. On the other hand, a study by Colombe et al. (2018) reported the seroprevalence of brucellosis and Q fever in small ruminants at a Thai- Cambodia border community at 13.3% and 33.3% respectively. Our surveillance did not collect small ruminant samples as these animals are commonly slaughtered at the household level or restaurants. A surveillance program of both diseases in other susceptible hosts like small ruminants would be required to better understand these zoonotic disease risks in the human population.
Interpretation of the CSF and PRRS serology results is difficult as these abattoir-collected samples had no vaccination history, and vaccination programs for large commercial pig farms in Cambodia were commercial-in-confidence. Vaccines against CSF and PRRS are commonly used in pig production in South East Asia (Kunavongkrit & Heard, 2000; Thammakarn, Hung, & Eardmusic, 2018; Zhang et al., 2017) and are widely available in Cambodia. Moreover, the diagnostic kits used in this study could not differentiate antibodies arising post-vaccination from those arising post-infection. High seroprevalences of CSF and PRRS detected in our study most likely resulted from vaccine-induced antibodies as more than 90% of the pig samples were from large commercial farms. Thai pigs showed higher seroprevalence than the Cambodian large commercial farm pigs. On the other hand, the seroprevalence of CSF and PRRS in pigs from the local small scale commercial farms may indicate some natural infection. CSF and PRRS vaccine use in smallholders and semi-commercial farms were reportedly low (Sothoeun, Young, & Windsor, 2013; Tornimbene et al., 2014; Tornimbene et al., 2015; Zhang et al., 2017). However, the total samples from local small to medium holders were really low in our study. Information on vaccination records and actual practices in commercial farms and smallholders would help explain the findings. The low numbers of smallholder pigs at abattoirs likely reflected population decline (due to competition from commercial piggeries, and perhaps ASF outbreaks) and cultural practices where these pigs were often slaughtered at the household level.
Risk factor analyses suggested that factors including sampling dates, sampling months, abattoirs location, and animal origin were potential predictors of sero-reactors. However, there is not enough information at this stage to conclude the role these factors play. The animal origin factor was not significantly correlated with the seropositive Q fever samples despite all four positive samples originated from Takeo. In the case of ASF tests, only the abattoir variable was significantly correlated to seropositive samples, not the animal origin. These observations could be due to the animal origin variable having a much larger denominator compared to the collection date for the Q fever dataset and to the abattoir for the ASF dataset. Further investigation in the areas where positive animals originated from would provide more in-depth disease information. The first confirmed ASF cases were reported in Ratanakiri province in April 2019, then five other provinces close to the Vietnam border were also confirmed with ASF outbreaks (FAO, 2020). Control measures implemented by the Cambodian government included movement controls on live pigs, and pig products and stamping out in the affected areas (FAO, 2020). However, scientific publications on ASF control in Cambodia were not available at the time of our study. All ASF antibody positive and doubtful samples were from three Cambodian large commercial farms and also from Thailand. There was no indication of ASF in either of these production sources – introduction of the currently circulating strain of ASF into commercial operations would be expected to result in high mortalities (FAO, 2020; Mazur-Panasiuk, Żmudzki, & Woźniakowski, 2019) and trigger some sort of alert, even if only in local media. There was no ASF outbreak reported in Cambodian origin provinces (Pig Progress, 2021a) during the time of our study, This raised then some concern as to the nature of the positive results. The Dse published by the ASF ELISA manufacturer’s internal validation was 95.8% while the Dsp was reported up to 100% (Dixon, 2014), indicating that false-negative, but not false-positive results could occur. All the positive samples were confirmed in a retest to rule out test aberration on the days of testing, suggesting that these animals had been previously exposed to the virus. Samples that tested positive and doubtful by the antibody ELISA technique were negative when tested in the RT-PCR, indicating they were not persistently infected survivors of infection. ASF is a highly contagious disease with a mortality rate of up to 100% and a vaccine was not available. The results would indicate that under field conditions the ASF ELISA has a Dse of less than 100%, unless there has been some use of vaccines that might have been imported from elsewhere. It is unlikely that an undetectable low virulence mutant (Pig Progress, 2021b) would emerge in Cambodia. Therefore, further investigations are required to determine the true nature of this ASF seropositivity.
In conclusion, an abattoir surveillance system could provide initial disease seroprevalence of high impact diseases and zoonoses for further investigation if needed. When resources are limited, the focuses of the survey should be adjusted based on the national priorities and current situations. The cost-effectiveness of a survey program could be increased through building provincial veterinary and para-veterinary capacities and cost-cutting where possible.