In western countries, the most frequent cause of human salmonellosis is inadequately processed, handled or cooked poultry products. The approaches used to control Salmonella in broiler chickens vary among regions and countries, and there is currently no mandated control programme for broiler production in Ontario or Canada, writes Dr Scott McEwen of the University of Guelph in his report for the Poultry Industry Council of Canada.

All steps of the transparent and replicable scoping study and structured risk were completed by two independent reviewers using an electronic structured risk format.

The majority of primary research focussed on Salmonella interventions (748 studies), followed by prevalence (200) and risk factors (30). Four on-farm interventions, including competitive exclusion (CE), other feed and water additives (with the exception of antimicrobials), vaccination and biosecurity, and two processing interventions, namely carcass spraying/dipping, and chilling, were prioritised for rigorous structured risk management approach (SR-MA).

The ranked list of interventions is: 1. CE was the most common on-farm intervention studied (192 studies); 2. Feed and water additives (other than antimicrobials) was the second most common on-farm intervention (114 studies) with 3. vaccination less common (69 studies). By contrast, a lack of primary research was observed for biosecurity practices, which are frequently recommended to producers for disease prevention and control. 4. At processing, the most commonly investigated intervention was carcass treatments with a spray or dip (75 studies) followed by 5. carcass chilling (37 studies).

Overall, the survey indicated that undefined CE products are more effective then defined products at reducing Salmonella colonisation in broilers, except for a defined, continuous-flow culture (CF3). A variety of routes, including as an additive to feed and water, were as effective at conferring protection as the most popular route studied, oral gavage. Researchers should further refine the development of defined products that are at least as effective as undefined products, and the specific characteristics that make CF3 products more effective than other defined CE products should be better understood.

Of the feed and water additives, prebiotics such as fructo-oligosaccharides, sucrose and lactose products showed the most promise against Salmonella infection in broilers.

Of the vaccines, a killed S. Typhimurium vaccine showed the most promise against Salmonella.

Of the biosecurity interventions, an electrostatic space charger in the hatchery also showed promise against Salmonella.

At the processing level, carcass treatments with a spray or dip as well as carcass chilling resulted in reductions of Salmonella prevalence and concentration. Trisodium phosphate (TSP) and lactic acid (LA) were the most commonly investigated carcass spray and dip treatments and were found to produce statistically significant reductions in Salmonella prevalence and concentration. However, the high pH of TSP and the undesirable changes in final product quality caused by lactic acid should be considered before these disinfectants are considered for use. Immersion chilling with chlorine was the most commonly investigated chilling treatment.

Results from the management approach indicated that chlorine is an effective chilling disinfectant, although the large amount of heterogeneity in the data limits the usefulness of an overall estimate of effectiveness. Rather, individual results should be reviewed to determine the range of possible reductions in Salmonella as a result of chilling with chlorine. The small number of studies investigating chilling methods other than immersion chilling with agitation limits conclusions that can be drawn from the data regarding the effectiveness of these chilling methods, although MAs indicate that overall, the different chilling types reduce the prevalence and concentration of Salmonella. The optimal combination of interventions is a blend of on-farm and processing interventions to reduce Salmonella from production to final processing.

Overall, the assessment of studies evaluating the effectiveness of various interventions – both at the farm and processing levels – has shown that much research is poorly conducted and/or reported.

The PIC findings support recent international initiatives advocating for the development and use of more stringent primary research reporting guidelines in animal health research.

Policy and decision-makers could use the results of this study for developing transparent, evidence-based guidelines for the on-farm Salmonella prevention and control in broiler chicken or to regulate the use of certain interventions (e.g. defined specific CE products).

There are some limitations that pertain to this study. Due to time and funding restrictions, only English language papers were included in the scoping review. Despite extensive efforts to uncover all existing primary research with thorough search verification, a statistical indication of publication bias was observed in the MA assessing the effectiveness of certain interventions, e.g. CE with studies reporting a concentration outcome.

It is possible that large, integrated poultry operations might have these data but keep them confidential (Sue Reynolds, Microbial Developments Ltd., Worcestershire, UK, personal communication) as they do not have incentive to publish trials, especially those with negative results.

Government and industry stakeholders could use the research synthesis framework proposed in this project as a standard tool for evaluating scientific evidence, particularly when the experts disagree and scientific recommendations are contradictory. Funding agencies could use this information to support new primary research targeting this important issue within the context of the knowledge gaps and future research needs that were identified through both scoping review and SR. The quantitative MA inputs would be used in our complementary QRA and comparison of prioritised interventions within the Canadian (Ontario) context.

Future research should evaluate the cost-effectiveness of various interventions addressed in this project.

November 2010