Food Protection Trends - January/February 2021 - 106

accurate, a tool needed by the food industry to effectively implement HACCP programs. CSPI learned a lot through this work
and importantly we shared it publicly with both industry and government agencies. We were in fact the first group to disclose that
fresh produce causes a significant number of illnesses and outbreaks.
In the early years of collecting and cataloguing this data,
state outbreak reports increased, and our database grew. By
CSPI  Foodborne  Outbreaks  Reported  to  CDC  2002--‐2011
2005, our outbreak database contained almost 4,500 outbreaks,
N=10,409	
  
including 138,000 individual cases of illness. CSPI continually
managed the outbreak database from 1997 to 2016 and developed a webpage to make the data available for free. When I left
CSPI to join the U.S. Food and Drug Administration (FDA) in
2015, the database contained more than 7,000 outbreaks with an
identified food and pathogen. But we also saw that state outbreak
reporting was flagging. Between 2002 and 2011, not only were
overall outbreak reports declining, so was the percent of those
reports that identified both the food and pathogen. Of the outbreaks reported to the CDC in 2011, for example, only 32% were
solved, which was a 10% decline from ten years earlier. That meant a lot of important information was being lost, information
that might help improve industry HACCP programs and help prevent the next outbreak.
To counter this, CSPI started analyzing state outbreak reporting and grading the states, based on their performance
over a 10-year period. While states known to have robust public health
CSPI  Grading  the  States:  Outbreak  ReporNng    
systems, like Minnesota and Oregon, reported six to eight outbreaks per
2002--‐2011
million
population, a large number of states reported fewer than one outbreak
*  A:	
  ≥8	
  outbreaks/
million	
  populaFon	
   per million, a consequence of weak state and local systems for managing
*  B:	
  6--‐7	
  outbreaks/
million	
  populaFon	
   public health and outbreak investigations. Please note, these same public
*  C:	
  4--‐5	
  outbreaks/
million	
  populaFon	
   systems are called on to manage the current COVID-19 pandemic. Perhaps
*  D:	
  2--‐3	
  outbreaks/
our research on foodborne outbreak reporting weaknesses foreshadowed the
million	
  populaFon	
  
*  F:	
  ≤1	
  outbreaks/
public health crisis we are currently facing.
million	
  populaFon	
  
Over the years of working with the outbreak data in the U.S., we
proved that this outbreak information was essential for the public to access,
not only for control of ongoing outbreaks, but to understand the trends over
time and where foods causing outbreaks were being consumed. It also gave us tools for evaluating federal and state agencies
and incentives for long-term surveillance improvements.
And now, I want to turn to newer innovations that I consider audacious and cutting-edge. The first examples are new
tools for outbreak surveillance. Imagine there is an app on your phone that
lets you record illnesses in your family or livestock in real time. Further imagine that the app links not only to this group of food safety specialists from
government, industry, and academia, but to clinical doctors, livestock veterinarians, and even wildlife experts. These apps are being pilot tested today.
One app, called AfyaData, tracked more than 13,000 reports in Tanzania
since its launch in 2015, with most reports coming from community sources.
This model for outbreak surveillance tracks real-time occurrences in a One
Health framework, so that outbreaks and epidemics in humans, livestock,
and even wildlife can be identified and isolated before they become pandemics.
This work was conducted by a group called Ending Pandemics, led by Mark Smolinski, which created the model for
community-led surveillance. An example in Thailand, called Participatory One Health Digital Disease Detection (PODD), is a
cell phone-based platform for community-based reporting of human, animal, and environmental health threats. Key to this vision
is having a set of metrics that can standardize reporting, including a set of Outbreak Timeliness Metrics. The metrics identify
key outbreak events and help public health officials measure and track the time between events, such as the start date of the
outbreak to the date of outbreak detection. The speed of such surveillance can correlate with the number of cases, severity of
outcomes, lives lost, and the cost to the public health systems. Approaches like crowd sourcing surveillance information could
expedite timely detection and response in countries and regions that currently lack surveillance systems, just like cell phones
eliminated the need for wiring entire countries for land lines.
The second example of bold innovation is the approach to information sharing developed by the Johns Hopkins University and the Global Alliance for Improved Nutrition, partnering with many others. The Food Systems Dashboard provides food
system data from multiple sources gathered and integrated on a platform that is accessible and easy to use. The Dashboard
provides data on Food Supply Chains, Food Environments, Individual Factors, Consumer Behavior and Diets, and provides
tools to compare that data by region and country, as well as over time. For example, looking at one country, you can access data
Solved	
  

Unsolved	
  

1,400	
  

Outbreaks	
  
	
  

1,200	
  
1,000	
  

800	
  

755	
  

640	
  

817	
  

400	
  
200	
  

809	
  

719	
  

667	
  

566	
  

454	
  

512	
  

424	
  

459	
  

379	
  

365	
  

237	
  

292	
  

245	
  

2002	
  

2003	
  

2004	
  

2005	
  

2006	
  

2007	
  

2008	
  

2009	
  

2010	
  

2011	
  

544	
  

600	
  

438	
  

0	
  

42.8%	
  Solved	
  

106

Food Protection Trends January/February

566	
  

521	
  

32.0%	
  Solved	
  

17
Food Protection Trends - January/February 2021

Table of Contents for the Digital Edition of Food Protection Trends - January/February 2021
