Risk of salmonellosis from consumption of almonds in the North American market

Salmonellosis outbreaks from consumption of raw almonds in 2001 and 2004 led to regulations that require mandatory treatment of almonds sold in North America to give a minimum 4-log reduction of Salmonella. This study aims to: 1) assess the risk of salmonellosis associated with almond consumption in North America, with current treatments in effect; 2) determine the resilience of the current production system to increases in prevalence or concentration of Salmonella on almonds; 3) assess the impact of treating less than 100% of the crop; and 4) investigate conditions that could explain the number of cases associated with the 2001 outbreak. Risk was assessed using a Monte Carlo simulation, based on an established dose–response relationship. Data for almond amounts sold, Salmonella prevalence and concentration on almonds, storage time and temperature at different handling steps, population reductions during storage at various temperatures and with different treatments, and consumer handling were based on data from published sources and almond industry or academic expert opinion. What-if scenarios were evaluated for Salmonella prevalence varying from 1 to 65%, concentrations of Salmonella varying from 1 to 120 MPN/100 g, and portions of untreated crop varying from 0 to 10%. The estimated incidence of salmonellosis in North America from almonds as currently treated is on average 0.008 cases per billion servings (with an estimated 6.6 billion servings consumed annually). Increases in Salmonella prevalence to 25%, mean concentrations above 25 MPN/100 g, or leaving 0.05% of the crop untreated all resulted in an arithmetic mean risk greater than 1 case/year (with geometric means remaining below 1 case/year for all variables). Assuming 4000 kg at a prevalence of 65% (observed in recalled lots) and an average concentration of 120 MPN/100 g in raw almonds (back calculated from levels in recalled almonds) predicted over the 2800 cases estimated for the 2001 outbreak. Applying a 4-log reduction to these almonds reduced the average number of predicted cases to less than a single case. The current regulation is effective in maintaining the risk of salmonellosis from consumption of almonds below an arithmetic mean of 1 case/year, although significant increases in either prevalence or concentration, or small increases in proportion of untreated almonds would frequently lead to exceeding this threshold.

► Low risk of salmonellosis predicted from consumption of processed almonds. ► Risk rises to > 1 arithmetic mean cases per year with moderate increases in either prevalence or concentration and small increases in percent cropuntreated. ► Salmonella prevalence and concentration in raw almonds atypical in 2001 outbreak. ► Model predicts a 4-log treatment could have prevented the 2001 outbreak.