Foodborne illness is prevented by inspection and surveillance conducted by health departments across America. Appropriate restaurant behavior is enforced and monitored via public health inspections. However, surveillance coverage provided by state and local health departments is insufficient in preventing the rising number of foodborne illness outbreaks. To address this need for improved surveillance coverage we conducted a supplementary form of public health surveillance using social media data: Yelp.com restaurant reviews in the city of San Francisco. Yelp is a social media site where users post reviews and rate restaurants they have personally visited. Presence of keywords related to health code regulations and foodborne illness symptoms, number of restaurant reviews, number of Yelp stars, and restaurant price range were included in a model predicting a restaurant’s likelihood of health code violation measured by the assigned San Francisco public health code rating. For a list of major health code violations see (S1 Table). We built the predictive model using 71,360 Yelp reviews of restaurants in the San Francisco Bay Area. The predictive model was able to predict health code violations in 78% of the restaurants receiving serious citations in our pilot study of 440 restaurants. Training and validation data sets each pulled data from 220 restaurants in San Francisco. Keyword analysis of free text within Yelp not only improved detection of high-risk restaurants, but it also served to identify specific risk factors related to health code violation. To further validate our model we applied the model generated in our pilot study to Yelp data from 1,542 restaurants in San Francisco. The model achieved 91% sensitivity 74% specificity, area under the receiver operator curve of 98%, and positive predictive value of 29% (given a substandard health code rating prevalence of 10%). When our model was applied to restaurant reviews in New York City we achieved 74% sensitivity, 54% specificity, area under the receiver operator curve of 77%, and positive predictive value of 25% (given a prevalence of 12%). Model accuracy improved when reviews ranked highest by Yelp were utilized. Our results indicate that public health surveillance can be improved by using social media data to identify restaurants at high risk for health code violation. Additionally, using highly ranked Yelp reviews improves predictive power and limits the number of reviews needed to generate prediction. Use of this approach as an adjunct to current risk ranking of restaurants prior to inspection may enhance detection of those restaurants participating in high risk practices that may have gone previously undetected. This model represents a step forward in the integration of social media into meaningful public health interventions.

Analysis of ‘Food Inspection’ provided by Analyst-2 (analyst-2.ai), based on source dataset retrieved from https://catalog.data.gov/dataset/5e5361d0-0bc9-4787-81ef-fee13b3a00b5 on 13 February 2022.

--- Dataset description provided by original source is as follows ---

The Licensure & Regulatory Services Program inspects all licensed retail food establishments in Montgomery County for a variety of reasons (e.g. obtaining a permit, regular check-ups, or in response to complaints.) Included in this overall surveillance are two types of inspections that are conducted on a routine basis. The first type, a comprehensive inspection, is a thorough inspection that evaluates the sanitation, maintenance, and food service operations for the facility. This inspection includes many of the items found in the monitoring inspection, such as critical temperatures and food handling procedures. The second type, a monitoring inspection, involves checking the food service operation for critical food temperatures, equipment temperatures, and general food handling / cleanliness practices. While monitoring inspections are not as detailed as environmental inspections, they help ensure that the facility is operating safely throughout the year.

The frequency of inspections is based on the food-borne illness risks associated with the food being processed at a facility. Based on the risk each facility is designated as high, medium, or low priority. High priority risk facilities prepare food products a day or more in advance or utilize any combination of two or more processes such as a cooking, cooling, reheating and hot holding food over four hours. Moderate priority risk facilities prepare and cook food products which are served immediately or within four hours. A low priority risk facility serve prepackage foods that are not potentially hazardous.

For high risk facilities, a monitoring inspection is conducted at least twice a year, and moderate risk facilities are inspected at least once a year. High and moderate risk facilities also receive at least one comprehensive inspection every year. Low risk priority facilities typically receive one comprehensive inspection every two years.

(C) Denotes a Critical Violation. A Critical Violation means a food safety requirement that requires immediate correction. Failure for immediate correction results in cessation of some or all food operations or closure of the facility until violation is able to be corrected.

Update Frequency: Daily

--- Original source retains full ownership of the source dataset ---

Here are a few use cases for this project:

1. Food Safety Inspection: The Bacteria Detection V3 model can be used in food processing plants and restaurants to inspect raw materials, cooked meals, and equipment surfaces for the presence of harmful bacteria, ensuring food safety and reducing the risk of foodborne illnesses.

2. Water Quality Monitoring: Municipal water treatment plants and environmental agencies can use the model to analyze water samples, detecting any contamination with harmful bacteria strains. This will help in maintaining safe drinking water standards and monitoring the health of aquatic ecosystems.

3. Medical Diagnostics: Healthcare professionals can utilize the Bacteria Detection V3 model to quickly and accurately identify bacterial infections in clinical samples, such as blood, urine, or swabs. This will aid in early diagnosis and targeted treatment, improving patient outcomes.

4. Pharmaceutical Manufacturing: The pharmaceutical industry can implement the model for quality control purposes, identifying potential bacterial contamination in drug manufacturing processes, and ensuring that the final product is safe for consumption.

5. Agricultural and Livestock Management: Farmers and livestock managers can use the model to detect harmful bacteria in their animals' feed, water, and living environments. This will help maintain healthy livestock, minimize the risk of disease outbreaks, and increase overall productivity.

As of 2021, the LIVES standard is no longer being used to evaluate restaurants. For that reason, this dataset is no longer being updated but will remain on the open data portal. For new inspection data please visit this website or download the new data here (Health Inspections (2023-Present)). SUMMARY This dataset contains the results of health inspections conducted by the Department of Public Health from 2016-2019. It includes the name and location of each facility inspected, the inspection score, and any violations observed. HOW THIS DATASET IS CREATED After inspecting a facility, a Health Inspector assigns a score based on observed violations. Violations are categorized by risk level: High risk category: records specific violations that directly relate to the transmission of food borne illnesses, the adulteration of food products and the contamination of food-contact surfaces. Moderate risk category: records specific violations that are of a moderate risk to the public health and safety. Low risk category: records violations that are low risk or have no immediate risk to the public health and safety. The score card that will be issued by the inspector is maintained at the food establishment and is available to the public in this dataset. UPDATE PROCESS This dataset is historical and will not be updated RELATED DATASETS Health Inspections (2023-Present) Health Inspections (2020-2023) Health Inspections (2016-2019) Inspection lookup tool

Context:

The SF Health Department has developed an inspection report and scoring system. After conducting an inspection of the facility, the Health Inspector calculates a score based on the violations observed. Violations can fall into:high risk category: records specific violations that directly relate to the transmission of food borne illnesses, the adulteration of food products and the contamination of food-contact surfaces.moderate risk category: records specific violations that are of a moderate risk to the public health and safety.low risk category: records violations that are low risk or have no immediate risk to the public health and safety.The score card that will be issued by the inspector is maintained at the food establishment and is available to the public in this dataset.

Potential question(s) to get started with!

• What are some predictors of health scores? What relevant outside data can you bring to bear on the question, including restaurant reviews, sentiment analysis, demographic data, etc?

Fields:

San Francisco's LIVES restaurant inspection data leverages the LIVES Flattened Schema (https://goo.gl/c3nNvr), which is based on LIVES version 2.0, cited on Yelp's website (http://www.yelp.com/healthscores).

Please refer to https://goo.gl/c3nNvr for detailed data dictionary.

Further info on the Food Safety Program can be found here.

We have included the following commonly used geographic shapefiles:

• Analysis Neighborhoods
• Supervisor Districts as of April 2012

Acknowledgements:

Data provided by the San Francisco Health Department via the San Francisco Open Data Portal at https://data.sfgov.org/d/pyih-qa8i License: PDDL 1.0 ODC Public Domain Dedication and Licence (PDDL)

Photo via Flickr Rob Hyndman Attribution-NonCommercial-ShareAlike 2.0 Generic (CC BY-NC-SA 2.0)

NNDSS - Table 1E. Botulism, Foodborne to Botulism, Other (wound & unspecified) - 2019. In this Table, provisional cases* of notifiable diseases are displayed for United States, U.S. territories, and Non-U.S. residents.

Note: This table contains provisional cases of national notifiable diseases from the National Notifiable Diseases Surveillance System (NNDSS). NNDSS data from the 50 states, New York City, the District of Columbia and the U.S. territories are collated and published weekly on the NNDSS Data and Statistics web page (https://wwwn.cdc.gov/nndss/data-and-statistics.html). Cases reported by state health departments to CDC for weekly publication are provisional because of the time needed to complete case follow-up. Therefore, numbers presented in later weeks may reflect changes made to these counts as additional information becomes available. The national surveillance case definitions used to define a case are available on the NNDSS web site at https://wwwn.cdc.gov/nndss/. Information about the weekly provisional data and guides to interpreting data are available at: https://wwwn.cdc.gov/nndss/infectious-tables.html

Footnotes: U: Unavailable — The reporting jurisdiction was unable to send the data to CDC or CDC was unable to process the data. -: No reported cases — The reporting jurisdiction did not submit any cases to CDC. N: Not reportable — The disease or condition was not reportable by law, statute, or regulation in the reporting jurisdiction. NN: Not nationally notifiable — This condition was not designated as being nationally notifiable. NP: Nationally notifiable but not published — CDC does not have data because of changes in how conditions are categorized. Cum: Cumulative year-to-date counts. Max: Maximum — Maximum case count during the previous 52 weeks. * Case counts for reporting years 2018 and 2019 are provisional and subject to change. Cases are assigned to the reporting jurisdiction submitting the case to NNDSS, if the case's country of usual residence is the US, a US territory, unknown, or null (i.e. country not reported); otherwise, the case is assigned to the 'Non-US Residents' category. For further information on interpretation of these data, see https://wwwn.cdc.gov/nndss/document/Users_guide_WONDER_tables_cleared_final.pdf. † Previous 52 week maximum and cumulative YTD are determined from periods of time when the condition was reportable in the jurisdiction (i.e., may be less than 52 weeks of data or incomplete YTD data).

Despite being the main source of fresh, convenient, and affordable food for 80% of Hanoi’s population, food flows within traditional markets remain largely invisible due to a lack of tracing systems and environmental conditions which make traditional tracking approaches challenging.

By providing free internet to a series of wholesalers and markets in the Cau Giay and Dong Anh districts of Hanoi, Vietnam, this project will put in place the first pieces of tracking system that will characterize and monitor food flows between traders, retailers, and consumers.

Research has found that 10-40% of traditional market food is contaminated with microbes or parasites which cause foodborne illnesses. As shoppers become increasingly concerned about food safety and large-scale retailers that can offer food safety certification expand rapidly, this project aims to equip traditional market actors with data that could prevent their marginalization through urban policy decisions that may favor organized retailers, as well as improve the safety of traditional market goods.

The collected food flow data will allow for improved linkages among key traditional market actors and help identify better policy and planning options for improving distribution channels in ways that benefits under-resourced communities.

To implement the project, the Alliance of Bioversity International and CIAT and the General Statistics Office (GSO) of Vietnam survey actors and track space and time data points on all devices within the range of the WiFi routers and signal amplifiers, whether connected to the internet or not.

The pilot system ran on three layers of data:

Layer One

Every smartphone has a unique media access control (MAC) address that the WiFi routers installed in the markers use to identify how many MAC addresses visit the markets over time, how many return to the market and how often, and how markets differ on these metrics. This data is collected even if the smartphone is not connected to the WiFi network.

Layer Two

When a smartphone user connects to the free WiFi, they are prompted to answer a series of questions depending on their user type (vendor, customer, etc.). For example, a user that identifies as a vendor is asked questions regarding sales of specific commodities which will allow for sales to be characterized across time and space.

Layer Three

To validate findings in Layer One and Two, in-person surveys were conducted with vegetable, pork and rice sellers in five traditional markets in Hanoi

mac: An anonymized version of the MAC. All the MAC address were anonymized through a SHA-3 256 hashing function. The hashed mac ensure anonymity while is consistent across all markets and during the whole period of the analysis. We can therefore ensure that a given mac found in two different dataset will correspond to the same phone.

market: The name of the market where the phone was seen

role: Self-identified role if the user connected to the wifi and filled-out the layer 2 form

gender: Self-identified role if the user connected to the wifi and filled-out the layer 2 form

median_first_seen: The median time when the user is first seen in the markets (in minutes starting at 0 from midnight) (e.g. the time the user entered the market)

median_last_seen: The median time when the user is last seen in the markets (in minutes starting at 0 from midnight) (e.g. the time the user left the market)

average_time_day: The average number of time the user visited the market. A period of time of at least 2 hours between two consecutive observation of the user in the market is needed to be counted as a different visit.

average_duration_day: The average duration spent on the market daily.

average_day_week: The average number of visits per week.

average_total_day_seen: The total number of days a user was seen on the market.

total_durantion: Total duration spent by a single user on the market.

Recent foodborne illness outbreaks have heightened pressures on growers to deter wildlife from farms, jeopardizing conservation efforts. However, it remains unclear which species, particularly birds, pose the greatest risk to food safety. Using >11,000 pathogen tests and 1,565 bird surveys covering 139 bird species from across the western U.S.A., we examined the importance of 11 traits in mediating wild bird risk to food safety. We tested whether traits associated with pathogen exposure (e.g., habitat associations, movement, and foraging strategy) and pace-of-life (clutch size and generation length) mediated foodborne pathogen prevalence and proclivities to enter farm fields and defecate on crops. Campylobacter spp. were the most prevalent enteric pathogen (8.0%), while Salmonella and Shiga-toxin producing E. coli (STEC) were rare (0.46% and 0.22% prevalence, respectively). We found that several traits related to pathogen exposure predicted pathogen prevalence. Specifically, Campylobacter and STEC-associated virulence genes were more often detected in species associated with cattle feedlots and bird feeders, respectively. Campylobacter was also more prevalent in species that consumed plants and had longer generation lengths. We found that species associated with feedlots were more likely to enter fields and defecate on crops. Our results indicated that canopy-foraging insectivores were less likely to deposit foodborne pathogens on crops, suggesting growers may be able to promote pest-eating birds and birds of conservation concern (e.g., via nest boxes) without necessarily compromising food safety. As such, promoting insectivorous birds may represent a win-win-win for bird conservation, crop production, and food safety. Collectively, our results suggest that separating crop production from livestock farming may be the best way to lower food safety risks from birds. More broadly, our trait-based framework suggests a path forward for co-managing wildlife conservation and food safety risks in farmland by providing a strategy for holistically evaluating the food safety risks of wild animals, including under-studied species.

About one third of foodborne illness outbreaks in Europe are acquired in the home and eating undercooked poultry is among consumption practices associated with illness. The aim of this study was to investigate whether actual and recommended practices for monitoring chicken doneness are safe. Seventy-five European households from five European countries were interviewed and videoed while cooking chicken in their private kitchens, including young single men, families with infants/in pregnancy and elderly over seventy years. A cross-national web-survey collected cooking practices for chicken from 3969 households. In a laboratory kitchen, chicken breast fillets were injected with cocktails of Salmonella and Campylobacter and cooked to core temperatures between 55 and 70 °C. Microbial survival in the core and surface of the meat were determined. In a parallel experiment, core colour, colour of juice and texture were recorded. Finally, a range of cooking thermometers from the consumer market ...

The Licensure & Regulatory Services Program inspects all licensed retail food establishments in Montgomery County for a variety of reasons (e.g. obtaining a permit, regular check-ups, or in response to complaints.) Included in this overall surveillance are two types of inspections that are conducted on a routine basis. The first type, a comprehensive inspection, is a thorough inspection that evaluates the sanitation, maintenance, and food service operations for the facility. This inspection includes many of the items found in the monitoring inspection, such as critical temperatures and food handling procedures. The second type, a monitoring inspection, involves checking the food service operation for critical food temperatures, equipment temperatures, and general food handling / cleanliness practices. While monitoring inspections are not as detailed as environmental inspections, they help ensure that the facility is operating safely throughout the year.

The frequency of inspections is based on the food-borne illness risks associated with the food being processed at a facility. Based on the risk each facility is designated as high, medium, or low priority. High priority risk facilities prepare food products a day or more in advance or utilize any combination of two or more processes such as a cooking, cooling, reheating and hot holding food over four hours. Moderate priority risk facilities prepare and cook food products which are served immediately or within four hours. A low priority risk facility serve prepackage foods that are not potentially hazardous.

For high risk facilities, a monitoring inspection is conducted at least twice a year, and moderate risk facilities are inspected at least once a year. High and moderate risk facilities also receive at least one comprehensive inspection every year. Low risk priority facilities typically receive one comprehensive inspection every two years.

(C) Denotes a Critical Violation. A Critical Violation means a food safety requirement that requires immediate correction. Failure for immediate correction results in cessation of some or all food operations or closure of the facility until violation is able to be corrected.

Update Frequency: Daily

According to Canada’s Food Report Card 2016, there are 4 million foodborne illnesses acquired each year in the nation alone. The leading causes of foodborne illness are pathogenic bacteria such as shigatoxigenic/verotoxigenic Escherichia coli (STEC/VTEC) and Listeria monocytogenes. Most current detection methods used to identify these bacterial pathogens are limited in their validity since they are not specific to detecting metabolically active organisms, potentially generating false-positive results from non-living or non-viable bacteria. Previously, our lab developed an optimized bioorthogonal non-canonical amino acid tagging (BONCAT) method which allows for the labeling of translationally active wild-type pathogenic bacteria. Incorporation of homopropargyl glycine (HPG) into the cellular surfaces of bacteria allows for protein tagging using the bioorthogonal alkyne handle to report on the presence of pathogenic bacteria. Here, we use proteomics to identify more than 400 proteins differentially detected by BONCAT between at least two of five different VTEC serotypes. These findings pave the way for future examination of these proteins as biomarkers in BONCAT-utilizing assays.

This dataset was produced as part of the project: Reducing plastic packaging and food waste through product innovation simulation, funded by the NERC (Ref: NE/V010654/1). The wider project was concerned with building a Discrete Event Simulation (DES) model that can simulate food and food packaging as it moves through a household, from purchase to disposal. This dataset is qualitative in nature and was created to inform the modelling process. Remote qualitative interviews were planned and conducted, and participants were also asked to fill out entries on a research diary covering 4 days within a 7 day period. The interviews were semi-structured, and questions relating to waste and food practices were planned around 4 main areas: general household background, daily routines, weekly routines and shopping habits, waste and disposal practices, and attitudes to potential innovation in food packaging. A fifth section was based on photo-elicitation, to gauge participant attitudes to ageing produce. The diaries provided to participants had a system of prompts and questions that aimed to capture the movement of particular food items through the home and practices related to this movement over four separate days. Image files could also be submitted by participants, either included in the text files or as email attachments to the researcher. Follow up interviews were planned, which would be conducted after the second diary entry was complete and returned, but time constraints meant that only one follow up interview was conducted. A screening questionnaire was used to recruit and select participants, which was disseminated through university research recruitment channels and through social media accounts set up for the project. Informed consent for this questionnaire was gained at that stage, but was also gained prior to the remote interview, for both the interview and the diary research. However, the questionnaire data was not analysed and is not included in the collection as it was for recruitment purposes only. The data collection is comprised of 28 interview transcripts, 25 completed research diaries and 91 image files (produced by research participants) as part of the research diaries. Where image files were included in text documents by participants, the image files have been extracted and saved as photographs separately. These are stored as .docx files and .tiff files accordingly. Each of these was anonymised according to the anonymization plan. As there are multiple forms of data for each participant, each participant that completed the questionnaire was given a unique identifier, with suffixes describing the type of data. These are described in the readme file for the data collection.

THE PROBLEM Plastic packaging waste is a major issue that has recently entered public consciousness, with the British government committing to a 25-year plan that would phase out disposable packaging by 2042. Around 41% of plastic packaging is used for food, with the UK generating 1 million tonnes per year of packaging waste. Food packaging has had a 1844% increase in recycling since 2007, yet still only one third of food packaging is currently recycled [3]. Currently many consumers are boycotting plastic packaging. However, this is leading to a rise in food waste (and foodborne illness risk) due to decreased shelf life. Up to a third of the resources used to produce food could be saved by eliminating food waste [1]. In the UK, approximately 10 million tonnes of food are wasted every year, with the average family (i.e. a household containing children) spending £700 a year on food that is wasted. 31% of avoidable household food waste (1.3 million tonnes), is caused by a mismatch of packaging, pack, and portion size, and household food habits [2]. Plastic pollution and food waste can be reduced through product re-design and other household interventions. However, there is little evidence to determine the best solutions to reduce plastic pollution and food waste. The food industry and consumers have a variety of possible solutions, but no way of knowing the impacts and unintended consequences (without costly, time consuming trials and measurement). This is a major barrier to empowering the food system to enable the rapid reduction of plastic waste.

THE VISION This project reduces plastic pollution (and food waste) by providing a decision support tool to trigger action in the food industry and by consumers. Evidence concerning plastic and food waste reduction (and trade-offs with cost, and environmental impacts) will be generated by updating the Household Simulation Model (HHSM). The HHSM was piloted by the University of Sheffield and WRAP (the Waste & Resources Action Programme) to model the impacts of food product innovation quickly, to enable manufacturers to select the best innovations and interventions, and to prioritise their development and deployment. This project will incorporate into the current HHSM, data on 1)...

The Outbreak Management Division (OMD) is notified of multi-jurisdictional foodborne and zoonotic illness outbreaks and clusters by various partners including international, federal, provincial and territorial public health and food safety organizations. Once OMD is notified, these outbreaks and clusters are assessed to determine the appropriate follow up action related to them. The target performance indicator for this work is to ensure 90% of multi-jurisdictional foodborne and zoonotic illness outbreaks and clusters that OMD is notified of are assessed within 24 hours.