Salmonella control scenarios

In order to assess the costs and benefits of the pig feed salmonella control program, the costs of Salmonella monitoring and preventive measures, as well as the costs of feedborne cases of salmonella in Finland were estimated. The current situation was parameterized as described in the previous sections. In the current situation, there are statutory requirements to prevent salmonella contamination, and operators also apply voluntary preventive measures. Although salmonella control in feeds covers measures only until the pig eats the feed, there are costs throughout the chain that must be taken into account. In others words, costs incurred by feedborne salmonella cases on farms, at slaughterhouse, and human infections must also be taken into account, as they are affected by the efficiency of salmonella control of feeds. In the

case of salmonella contamination in the pork supply chain, the current regulation requires that the pathogen is eradicated immediately.

The estimated total costs of salmonella monitoring and prevention consist of the costs of preventive measures and the costs due to salmonella outbreaks and contaminations. The costs of preventive measures were calculated as the average costs per year in Finland and were as reported in Table 17. The costs of salmonella outbreaks and contaminations, should they occur, were simulated by Monte Carlo simulation model programmed in Matlab R2014b 8.4.0.150421 (Mathworks Inc., USA). All parameters (mean, median, standard deviation, percentiles) describing the prevalence of salmonella in different stages of the supply chain were obtained from the risk assessment scenarios presented in section 5.5 of this report. The costs associated with each contamination were determined as described in sections 7.5 and 7.6 of this report.

The costs of the current salmonella monitoring and control policy were compared with an alternative scenario. Hence, the two main scenarios analyzed were:

■ The current pig feed salmonella control program

■ A situation where the salmonella prevalence is similar to that presented in section 5.5.4. “Scenario 5: Salmonella prevalence in compound feed increases”.

In other words, the salmonella prevalence in pig feeds is assumed to be similar to the “EU average”. This outcome is assumed to be obtained by the current pig feed salmonella control not being applied by the commercial feed manufacturers (feed mill and mobile mixers), importers, and other operators. This is referred to as scenario A.

Alternative scenarios A assumed a lower level of salmonella monitoring and prevention when the contamination of feed is concerned (Table 25), and they therefore incurred lower prevention and monitoring costs than the current program. In the alternative scenario, no measures were assumed to be taken to eradicate salmonella when feed is contaminated, but in the event that salmonella occurred in pigs or humans, the same measures were assumed to be taken as in the current situation, because these measures are not part of the pig feed salmonella control program.

Cost-benefit analysis scenario A assumes that regarding preventive measures, the statutory measures required by the control program are not required. For example, salmonella sampling (either official by authorities or by self-control) were on a lower level (10–50% below the current level). However, the majority of the control measures were assumed to be realized (Table 26). The share of measures that would be applied in scenario A was determined after consulting four experts in the feed sector. Reduced control measures were expected to increase the prevalence of salmonella. Cost-benefit analysis scenario A corresponded to the Scenario 5 in section 5.5.4: In the worst scenario, all Finnish pigs would eat this feed, and the increase could be as high as 55–fold (10- to 130-fold).

Table 25. Summary of feedborne salmonella control measures applied in the current situation and in scenario A.

Summary of feedborne salmonella control measures applied in the current situation

and in cost-benefit analysis scenario A. Current situation Scenario A¹ Import and storage

Official control Yes No

Self-monitoring (sampling) Yes Reduced

Quarantine storage Yes No

Eradication if Salmonella detected Yes No

Commercial feed manufacturing

Treatment Yes Reduced

Hygiene measures Yes Slightly reduced

Pest control Yes Slightly reduced

Official control / sampling Yes No

Samples as self-control Yes Reduced

Self-monitoring (documentation) Yes Reduced

Eradication if Salmonella detected² Yes No

Mobile mixers

Treatment Yes Reduced

Official control / sampling Yes No

Hygiene measures Yes Yes

Samples as self-control Yes Reduced

Eradication if Salmonella detected Yes Reduced

Pig farms

Treatment (acid) Yes Yes

Sampling Yes Yes

Hygiene measure^s Yes Yes

Eradication if Salmonella detected Yes Yes, if in pigs Slaughterhouses

Extra measures if salmonella is detected Yes Yes

1 Scenario A corresponds to the worst-case scenario in section “5.5.4 Scenario 5: Salmonella prevalence in compound feed increases”. In the worst scenario, all Finnish pigs would eat this feed, and the increase could be as high as 55-fold (10- to 130-fold).

2 No costs of eradication accounted if salmonella detected in feed. This practice (no eradication if salmonella detected) does not correspond to the EU practice for compound feed.

Table 26. Share of preventive and monitoring measures applied in scenario A by feed business operators when compared to their application rate (100%) in the current situation.

Scenario A Commercial feed manufacturers

Share of measures still operated Feed mill operators Mobile mixers

Sampling as self-monitoring 50–90% 50–90%

Treatment (time and materials) 95 % 50 %

Treatment (installation and maintenance) 100 %

Official control 0 % 0 %

Sampling by official control 0 % 0 %

Hygiene measures 100 % 100 %

Pest control 100 %

Time used for other self-monitoring measures

(documenting measures etc.) 80 % 75 %

In addition, a sensitivity analysis was conducted regarding the current scenario. In the sensitivity analysis scenarios, the prevalence of feedborne salmonella at different stages of the production chain were increased by using an exponential scale. Hence, in the Exp 1 scenario, the incidence was 2.7-fold higher, in the Exp 2 scenario 7.4-fold higher, in the Exp 3 scenario 20.1-7.4-fold higher, and in the Exp 4 scenario 54.6-fold higher when compared to the 2013 prevalence. In sensitivity analysis, the prevalence of salmonella at different stages of the production chain was increased systematically. The sensitivity analysis scenarios only focused on the costs of realized salmonella contaminations. Salmonella monitoring was assumed to continue as it is currently, and only the prevalence and the costs of contaminations were assumed to change when compared to the current situation whereas preventive measures were not examined in these four scenarios.