Volume 27, Number 1—January 2021
Research
Recency-Weighted Statistical Modeling Approach to Attribute Illnesses Caused by 4 Pathogens to Food Sources Using Outbreak Data, United States
, LaTonia C. Richardson, Michael C. Bazaco, Cary Chen Parker, Stuart J. Chirtel, Dana Cole1, Neal J. Golden, Patricia M. Griffin, Weidong Gu2, Susan K. Schmitt3, Beverly J. Wolpert, Joanna S. Zablotsky Kufel, and R. Michael Hoekstra4
Table 2
Estimated percentages of foodborne illnesses attributed to 17 food categories and 90% credibility intervals for Salmonella, Escherichia coli O157, Listeria monocytogenes, and Campylobacter, based on analysis of single pathogen, single food category outbreaks, Foodborne Disease Outbreak Surveillance System, 1998–2012*
| Food category | % (90% credibility interval) |
|||
|---|---|---|---|---|
| Salmonella | E. coli O157 | L. monocytogenes | Campylobacter | |
| Land animals | ||||
| Beef | 9 (6–13) | 46 (36–55) | 0 (0–1) | 1 (<1–1) |
| Pork | 8 (6–10) | – | 2 (<1–8) | 3 (<1–8) |
| Chicken | 10 (7–13) | 0 (0–1) | 0 (0–2) | 8 (5–12) |
| Turkey | 7 (5–10) | 0 (0–<1) | 6 (2–16) | 2 (1–4) |
| Other meat or poultry | 0 (<1–1) | 0 (0–1) | – | 1 (<1–1) |
| Game | 0 (0–<1) | 1 (<1–3) | – | 0 (0–<1) |
| Dairy | 3 (1–5) | 9 (5–14) | 31 (12–64) | 66 (57–74) |
| Eggs |
12 (9–17) |
– |
– |
– |
| Aquatic animals | ||||
| Fish | 2 (1–3) | – | – | 0 (0–<1) |
| Other seafood |
0 (0–<1) |
– |
– |
6 (2–11) |
| Plants | ||||
| Grains, beans | 1 (<1–2) | – | – | – |
| Oils, sugars | – | – | – | 0 (0–1) |
| Fruits | 12 (8–16) | 7 (3–12) | 50 (5–77) | 1 (<1–2) |
| Seeded vegetables | 18 (13–25) | – | – | 6 (1–13) |
| Sprouts | 8 (5–12) | 1 (<1–1) | 8 (1–22) | – |
| Vegetable row crops | 3 (1–6) | 36 (26–46) | 3 (<1–13) | 6 (2–11) |
| Other produce | 7 (3–11) | 1 (0–2) | – | 2 (<1–6) |
*Estimates calculated by using analysis of variance model–estimated outbreak illnesses for single pathogen, single food category outbreaks occurring during 1998–2012, with down-weighting of outbreaks that occurred during 1998–2007. Because of rounding, 0 indicates nonzero estimates <0.5. Dashes indicate pathogen–food category pairs for which we did not estimate attribution percentages because of a lack of outbreaks.
1Current affiliation: US Department of Agriculture, Fort Collins, Colorado, USA.
2Current affiliation: US Department of Defense, Bethesda, Maryland, USA.
3Current affiliation: US Department of Veterans Affairs, Palo Alto, California, USA.
4Retired.