Making Better Use of Health Care Data

Curated on Posted on by Stefaan Verhulst

Benson S. Hsu, MD and Emily Griese in Harvard Business Review: “At Sanford Health, a $4.5 billion rural integrated health care system, we deliver care to over 2.5 million people in 300 communities across 250,000 square miles. In the process, we collect and store vast quantities of patient data – everything from admission, diagnostic, treatment and discharge data to online interactions between patients and providers, as well as data on providers themselves. All this data clearly represents a rich resource with the potential to improve care, but until recently was underutilized. The question was, how best to leverage it.

While we have a mature data infrastructure including a centralized data and analytics team, a standalone virtual data warehouse linking all data silos, and strict enterprise-wide data governance, we reasoned that the best way forward would be to collaborate with other institutions that had additional and complementary data capabilities and expertise.

We reached out to potential academic partners who were leading the way in data science, from university departments of math, science, and computer informatics to business and medical schools and invited them to collaborate with us on projects that could improve health care quality and lower costs. In exchange, Sanford created contracts that gave these partners access to data whose use had previously been constrained by concerns about data privacy and competitive-use agreements. With this access, academic partners are advancing their own research while providing real-world insights into care delivery.

The resulting Sanford Data Collaborative, now in its second year, has attracted regional and national partners and is already beginning to deliver data-driven innovations that are improving care delivery, patient engagement, and care access. Here we describe three that hold particular promise.

  • Developing Prescriptive Algorithms…
  • Augmenting Patient Engagement…
  • Improving Access to Care…(More)”.

The world’s first blockchain-powered elections just happened in Sierra Leone

Curated on Posted on by Stefaan Verhulst

Yomi Kazeem in Quartz: “On Mar. 7, elections in Sierra Leone marked a global landmark: the world’s first ever blockchain-powered presidential elections….

In Sierra Leone’s Western District, the most populous in the country, votes cast were manually recorded by Agora, a Swiss foundation offering digital voting solutions, using a permissioned blockchain. The idea was simple: just like blockchain technology helps ensure transparency with crytpocurrency transactions using public ledgers, by recording each vote on blockchain, Agora ensured transparency with votes cast in the district. While entries on permissioned blockchains can be viewed by everyone, entries can only be validated by authorized persons.

A lack of transparency has plagued many elections around the world, but particularly in some African countries where large sections of the electorate are often suspicions incumbent parties or ethnic loyalties have been responsible for the manipulation of the results in favor of one candidate or another. These suspicions remain even when there is little evidence of manipulation. A more transparent system could help restore trust.

Leonardo Gammar, CEO of Agora, says Sierra Leone’s NEC was “open minded” about the potential of blockchain in its elections after talks began late last year. “I also thought that if we can do it in Sierra Leone, we can do it everywhere else,” he says. That thinking is rooted in Sierra Leone’s developmental challenges which make electoral transparency difficult: poor network connectivity, low literacy levels and frequent electoral violence.

The big picture for Agora is to deploy solutions to automate the entire electoral process with citizens voting electronically using biometric data and personalized cryptographic keys and the votes in turn validated by blockchain. Gammar hopes Agora can replicate its work in other African elections on a larger scale but admits that doing so will require understanding the differing challenges each country faces.

Gammar says blockchain-powered electronic voting will be cheaper for African countries by cutting out the printing cost of paper-based elections but perhaps, more importantly, vastly reduce electoral violence…(More)”.

How to Make A.I. That’s Good for People

Curated on Posted on by Stefaan Verhulst

Fei-Fei Li in the New York Times: “For a field that was not well known outside of academia a decade ago, artificial intelligence has grown dizzyingly fast. Tech companies from Silicon Valley to Beijing are betting everything on it, venture capitalists are pouring billions into research and development, and start-ups are being created on what seems like a daily basis. If our era is the next Industrial Revolution, as many claim, A.I. is surely one of its driving forces.

It is an especially exciting time for a researcher like me. When I was a graduate student in computer science in the early 2000s, computers were barely able to detect sharp edges in photographs, let alone recognize something as loosely defined as a human face. But thanks to the growth of big data, advances in algorithms like neural networks and an abundance of powerful computer hardware, something momentous has occurred: A.I. has gone from an academic niche to the leading differentiator in a wide range of industries, including manufacturing, health care, transportation and retail.

I worry, however, that enthusiasm for A.I. is preventing us from reckoning with its looming effects on society. Despite its name, there is nothing “artificial” about this technology — it is made by humans, intended to behave like humans and affects humans. So if we want it to play a positive role in tomorrow’s world, it must be guided by human concerns.

I call this approach “human-centered A.I.” It consists of three goals that can help responsibly guide the development of intelligent machines.

First, A.I. needs to reflect more of the depth that characterizes our own intelligence….

No technology is more reflective of its creators than A.I. It has been said that there are no “machine” values at all, in fact; machine values arehuman values. A human-centered approach to A.I. means these machines don’t have to be our competitors, but partners in securing our well-being. However autonomous our technology becomes, its impact on the world — for better or worse — will always be our responsibility….(More).

How tech used to track the flu could change the game for public health response

Curated on Posted on by Stefaan Verhulst

Cathie Anderson in the Sacramento Bee: “Tech entrepreneurs and academic researchers are tracking the spread of flu in real-time, collecting data from social media and internet-connected devices that show startling accuracy when compared against surveillance data that public health officials don’t report until a week or two later….

Smart devices and mobile apps have the potential to reshape public health alerts and responses,…, for instance, the staff of smart thermometer maker Kinsa were receiving temperature readings that augured the surge of flu patients in emergency rooms there.

Kinsa thermometers are part of the movement toward the Internet of Things – devices that automatically transmit information to a database. No personal information is shared, unless users decide to input information such as age and gender. Using data from more than 1 million devices in U.S. homes, the staff is able to track fever as it hits and use an algorithm to estimate impact for a broader population….

Computational researcher Aaron Miller worked with an epidemiological team at the University of Iowa to assess the feasibility of using Kinsa data to forecast the spread of flu. He said the team first built a model using surveillance data from the CDC and used it to forecast the spread of influenza. Then the team created a model where they integrated the data from Kinsa along with that from the CDC.

“We got predictions that were … 10 to 50 percent better at predicting the spread of flu than when we used CDC data alone,” Miller said. “Potentially, in the future, if you had granular information from the devices and you had enough information, you could imagine doing analysis on a really local level to inform things like school closings.”

While Kinsa uses readings taken in homes, academic researchers and companies such as sickweather.com are using crowdsourcing from social media networks to provide information on the spread of flu. Siddharth Shah, a transformational health industry analyst at Frost & Sullivan, pointed to an award-winning international study led by researchers at Northeastern University that tracked flu through Twitter posts and other key parameters of flu.

When compared with official influenza surveillance systems, the researchers said, the model accurately forecast the evolution of influenza up to six weeks in advance, much earlier than prior models. Such advance warnings would give health agencies significantly more time to expand upon medical resources or to alert the public to measures they can take to prevent transmission of the disease….

For now, Shah said, technology will probably only augment or complement traditional public data streams. However, he added, innovations already are changing how diseases are tracked. Chronic disease management, for instance, is going digital with devices such as Omada health that helps people with Type 2 diabetes better manage health challenges and Noom, a mobile app that helps people stop dieting and instead work toward true lifestyle change….(More).

How We Identified Burned Villages in the Democratic Republic of Congo

Curated on Posted on by Stefaan Verhulst

Christophe Koettl in the New York Times: “In mid-February a source in the human rights community told me that villages in a remote region of the Democratic Republic of Congo were being burned amid a renewal of communal fighting. People fleeing the violence told aid workers of arson attacks.

The clashes between the Hema and Lendu communities — on the eastern side of the Ituri province, bordering Uganda — started in December and escalated in early February.

Historically, these distant conflicts have been difficult to analyze. But new technologies allow us to investigate them in close to real time.

I immediately collected active-fire data from NASA — thermal anomalies, or hot spots, that are recorded daily. It showed dozens of fires on the densely forested mountain ridge and along the shoreline of Lake Albert, one of the African Great Lakes between Congo and Uganda.

(Human rights groups also used this type of data, in combination with other evidence, to document the military’s scorched-earth campaign against the Rohingya in Myanmar.)

Active-fire data does not provide the cause of a fire, so one must exercise caution in interpreting it, especially when researching violence. It is more commonly used to track wildfires and agricultural fires.

The satellites that collect this information do not provide actual images; they only record the location of active fires, and very large ones at that. So don’t get your hopes up about watching your neighbors barbecue from space — we aren’t quite there yet.

Google and other online mapping platforms often show only blurry satellite images, or have no location names for remote areas such as the small fishing villages around Lake Albert. This makes it difficult to find places where people live. To deal with this challenge, I exported residential data from the online mapping site Openstreetmap.

I then overlaid the NASA data with this new data in Google Earth to look for recorded fires that were in or near populated places. This process gave me a shortlist of 10 locations to investigate.

Location of satellite-recorded active fires (the flames) and residential area data (the white outlines) helped to identify remote locations that had possibly been burned. © Google Earth/DigitalGlobe

Next, the satellite company DigitalGlobe provided me with high-resolution satellite imagery and analysis of these places. The results were disturbing: All the villages I had identified were at least partially burned, with hundreds of destroyed homes.

As this was not a comprehensive analysis of the whole area affected by violence, the actual number of burned villages is probably much higher. Aid organizations are reporting around 70 burned villages and more than 2,000 destroyed homes.

This new visual evidence provided us with a strong basis to report out the whole story. We now had details from both sides of the lake, not just at the refugee landing site in Uganda….(More)”

Exploring the Motives of Citizen Reporting Engagement: Self-Concern and Other-Orientation

Curated on Posted on by Stefaan Verhulst

Paper by Gabriel Abu-Tayeh, Oliver Neumann and Matthias Stuermer: “In smart city contexts, voluntary citizen reporting can be a particularly valuable source of information for local authorities. A key question in this regard is what motivates citizens to contribute their data. Drawing on motivation research in social psychology, the paper examines the question of whether self-concern or other-orientation is a stronger driver of citizen reporting engagement.

To test their hypotheses, the authors rely on a sample of users from the mobile application “Zurich as good as new” in Switzerland, which enables citizens to report damages in and other issues with the city’s infrastructure. Data was collected from two different sources: motivation was assessed in an online user survey (n = 650), whereas citizen reporting engagement was measured by the number of reports per user from real platform-use data. The analysis was carried out using negative binomial regression.

The findings suggest that both self-concern and other-orientation are significant drivers of citizen reporting engagement, although the effect of self-concern appears to be stronger in comparison. As such, this study contributes to a better understanding of what motivates citizens to participate in citizen reporting platforms, which are a cornerstone application in many smart cities….(More)”.

Data Science Landscape

Curated on Posted on by Stefaan Verhulst

Book edited by Usha Mujoo Munshi and Neeta Verma: “The edited volume deals with different contours of data science with special reference to data management for the research innovation landscape. The data is becoming pervasive in all spheres of human, economic and development activity. In this context, it is important to take stock of what is being done in the data management area and begin to prioritize, consider and formulate adoption of a formal data management system including citation protocols for use by research communities in different disciplines and also address various technical research issues. The volume, thus, focuses on some of these issues drawing typical examples from various domains….

In all, there are 21 chapters (with 21st Chapter addressing four different core aspects) written by eminent researchers in the field which deal with key issues of S&T, institutional, financial, sustainability, legal, IPR, data protocols, community norms and others, that need attention related to data management practices and protocols, coordinate area activities, and promote common practices and standards of the research community globally. In addition to the aspects touched above, the national / international perspectives of data and its various contours have also been portrayed through case studies in this volume. …(More)”.

Artificial intelligence could identify gang crimes—and ignite an ethical firestorm

Curated on Posted on by Stefaan Verhulst

Matthew Hutson at Science: “When someone roughs up a pedestrian, robs a store, or kills in cold blood, police want to know whether the perpetrator was a gang member: Do they need to send in a special enforcement team? Should they expect a crime in retaliation? Now, a new algorithm is trying to automate the process of identifying gang crimes. But some scientists warn that far from reducing gang violence, the program could do the opposite by eroding trust in communities, or it could brand innocent people as gang members.

That has created some tensions. At a presentation of the new program this month, one audience member grew so upset he stormed out of the talk, and some of the creators of the program have been tight-lipped about how it could be used….

For years, scientists have been using computer algorithms to map criminal networks, or to guess where and when future crimes might take place, a practice known as predictive policing. But little work has been done on labeling past crimes as gang-related.

In the new work, researchers developed a system that can identify a crime as gang-related based on only four pieces of information: the primary weapon, the number of suspects, and the neighborhood and location (such as an alley or street corner) where the crime took place. Such analytics, which can help characterize crimes before they’re fully investigated, could change how police respond, says Doug Haubert, city prosecutor for Long Beach, California, who has authored strategies on gang prevention.

To classify crimes, the researchers invented something called a partially generative neural network. A neural network is made of layers of small computing elements that process data in a way reminiscent of the brain’s neurons. A form of machine learning, it improves based on feedback—whether its judgments were right. In this case, researchers trained their algorithm using data from the Los Angeles Police Department (LAPD) in California from 2014 to 2016 on more than 50,000 gang-related and non–gang-related homicides, aggravated assaults, and robberies.

The researchers then tested their algorithm on another set of LAPD data. The network was “partially generative,” because even when it did not receive an officer’s narrative summary of a crime, it could use the four factors noted above to fill in that missing information and then use all the pieces to infer whether a crime was gang-related. Compared with a stripped-down version of the network that didn’t use this novel approach, the partially generative algorithm reduced errors by close to 30%, the team reported at the Artificial Intelligence, Ethics, and Society (AIES) conference this month in New Orleans, Louisiana. The researchers have not yet tested their algorithm’s accuracy against trained officers.

It’s an “interesting paper,” says Pete Burnap, a computer scientist at Cardiff University who has studied crime data. But although the predictions could be useful, it’s possible they would be no better than officers’ intuitions, he says. Haubert agrees, but he says that having the assistance of data modeling could sometimes produce “better and faster results.” Such analytics, he says, “would be especially useful in large urban areas where a lot of data is available.”…(More).

Infection forecasts powered by big data

Curated on Posted on by Stefaan Verhulst

Michael Eisenstein at Nature: “…The good news is that the present era of widespread access to the Internet and digital health has created a rich reservoir of valuable data for researchers to dive into….By harvesting and combining these streams of big data with conventional ways of monitoring infectious diseases, the public-health community could gain fresh powers to catch and curb emerging outbreaks before they rage out of control.

Going viral

Data scientists at Google were the first to make a major splash using data gathered online to track infectious diseases. The Google Flu Trends algorithm, launched in November 2008, combed through hundreds of billions of users’ queries on the popular search engine to look for small increases in flu-related terms such as symptoms or vaccine availability. Initial data suggested that Google Flu Trends could accurately map the incidence of flu with a lag of roughly one day. “It was a very exciting use of these data for the purpose of public health,” says Brownstein. “It really did start a whole revolution and new field of work in query data.”

Unfortunately, Google Flu Trends faltered when it mattered the most, completely missing the onset in April 2009 of the H1N1 pandemic. The algorithm also ran into trouble later on in the pandemic. It had been trained against seasonal fluctuations of flu, says Viboud, but people’s behaviour changed in the wake of panic fuelled by media reports — and that threw off Google’s data. …

Nevertheless, its work with Internet usage data was inspirational for infectious-disease researchers. A subsequent study from a team led by Cecilia Marques-Toledo at the Federal University of Minas Gerais in Belo Horizonte, Brazil, used Twitter to get high-resolution data on the spread of dengue fever in the country. The researchers could quickly map new cases to specific cities and even predict where the disease might spread to next (C. A. Marques-Toledo et al. PLoS Negl. Trop. Dis. 11, e0005729; 2017). Similarly, Brownstein and his colleagues were able to use search data from Google and Twitter to project the spread of Zika virus in Latin America several weeks before formal outbreak declarations were made by public-health officials. Both Internet services are used widely, which makes them data-rich resources. But they are also proprietary systems for which access to data is controlled by a third party; for that reason, Generous and his colleagues have opted instead to make use of search data from Wikipedia, which is open source. “You can get the access logs, and how many people are viewing articles, which serves as a pretty good proxy for search interest,” he says.

However, the problems that sank Google Flu Trends still exist….Additionally, online activity differs for infectious conditions with a social stigma such as syphilis or AIDS, because people who are or might be affected are more likely to be concerned about privacy. Appropriate search-term selection is essential: Generous notes that initial attempts to track flu on Twitter were confounded by irrelevant tweets about ‘Bieber fever’ — a decidedly non-fatal condition affecting fans of Canadian pop star Justin Bieber.

Alternatively, researchers can go straight to the source — by using smartphone apps to ask people directly about their health. Brownstein’s team has partnered with the Skoll Global Threats Fund to develop an app called Flu Near You, through which users can voluntarily report symptoms of infection and other information. “You get more detailed demographics about age and gender and vaccination status — things that you can’t get from other sources,” says Brownstein. Ten European Union member states are involved in a similar surveillance programme known as Influenzanet, which has generally maintained 30,000–40,000 active users for seven consecutive flu seasons. These voluntary reporting systems are particularly useful for diseases such as flu, for which many people do not bother going to the doctor — although it can be hard to persuade people to participate for no immediate benefit, says Brownstein. “But we still get a good signal from the people that are willing to be a part of this.”…(More)”.

Launching the Data Culture Project

Curated on Posted on by Stefaan Verhulst

New project by MIT Center for Civic Media and the Engagement Lab@Emerson College: “Learning to work with data is like learning a new language — immersing yourself in the culture is the best way to do it. For some individuals, this means jumping into tools like Excel, Tableau, programming, or R Studio. But what does this mean for a group of people that work together? We often talk about data literacy as if it’s an individual capacity, but what about data literacy for a community? How does an organization learn how to work with data?

About a year ago we (Rahul Bhargava and Catherine D’Ignazio) found that more and more users of our DataBasic.io suite of tools and activities were asking this question — online and in workshops. In response, with support from the Stanford Center on Philanthropy and Civil Society, we’ve worked together with 25 organizations to create the Data Culture Project. We’re happy to launch it publicly today! Visit datacultureproject.org to learn more.

The Data Culture Project is a hands-on learning program to kickstart a data culture within your organization. We provide facilitation videos to help you run creative introductions to get people across your organization talking to each other — from IT to marketing to programs to evaluation. These are not boring spreadsheet trainings! Try running our fun activities — one per month works as a brown bag lunch to focus people on a common learning goal. For example, “Sketch a Story” brings people together around basic concepts of quantitative text analysis and visual storytelling. “Asking Good Questions” introduces principles of exploratory data analysis in a fun environment. What’s more, you can use the sample data that we provide, or you can integrate your organization’s data as the topic of conversation and learning….(More)”.