I wasn’t so sure how to describe Treato. It’s really interesting – mixing semantic analysis, social media, medical records, and analysis – Treato is really the next level of how consumers research diseases.
The image to the right is one that came from searching for Lyme disease.
I am not sure how they assemble that graphic, but I guess it’s a mixture of curation (plainly telling the system the usual drugs) and seeing what folks report they took (by combing the web). [Aside: It seems to be just listing the most used drugs (and knowing how the disease is treated, it looks like that to me). I wonder if there's enough data for them to actually rank the drugs based on favorable outcomes, too. Outcomes analytics is where my head is always at, these days.]
Treato says they identified “over 160,000 English language websites, containing user-generated-content about medications and symptoms.” And that they “have searched and mapped thousands of these sites and indexed over 1 billion posts, covering well over 11,000 drug brands.” And they “have organized and analyzed this vast amount of data to create statistically proven insights about every aspect of medication use” [from their About page.]
I am not sure if they do this, but I’d like to see Treato applying their engine to sites like Patients Like Me, or 23 and Me, where there is an even richer set of data connecting people, disease, and outcomes.
And I am not sure, but they say the throw in patient records from HMOs, which I guess must mean claims data (which one can buy). But it would be cool if they could also mine medical record notes as well. They have the tech to do it, it’s just a matter of pointing their analytics tools to a new source (hm, what about Practice Fusion?).
I’m always thinking of how to extract meaning from medical records. So, Treato blows my mind.
Do you know of any other consumer disease research tools as sophisticated as Treato?
Hm, I wonder if they have enough to model trends, `a la Google Flu Trends, and then predict out-breaks? And it would be cool if Treato published a Health of the Nation – sort of like a Google Zeitgeist – of what is searched for and trends in diseases. I don’t think I’ve seen anything like that. Perhaps the CDC has something like this? For sure Treato is sitting on meaningful data they could use for this report. What do you think?
When I was at SXSW, I spent most of my time at either Big Data or Healthcare sessions. The interesting thing is that the Big Data sessions always devolved to talking about healthcare and the healthcare sessions always spoke about Big Data.
One aspect of the fusion of big data and healthcare that kept coming up was personal health monitoring. And since SXSW I have spoken and listened to a few folks (@dpatil, @rodrigoATCG, @nagle5000, @syntheticzero, and Sanjiv Shah) exploring measurement, self-analysis, and health devices.
Back in 2004-05, I worked on a product called Lifeblog – it collected all your SMS, MMS, photos, and videos into a timeline on your PC. It was way advanced for its time, and pointed to a future where we would cache our life’s stream of data. Working on that set me off thinking of sensors, lifestreaming, visualization of personal data (one set of discussions around visualizing recorded mobile phone usage data involved a colleague who left Nokia to set up @futureful – yes Nokia dabbled in this space once), my vision for what was supposed to be Ovi, my interest in lifestream aggregation, and a dotted line all the way to me getting involved in Big Data and Healthcare and Life Sciences in my role at IBM.
So, what do I see today in terms of these personal health devices?
At HIMSS in February, a big healthcare IT event, there was a whole section promoted by Qualcomm that was about mobile health measurement devices, such as the Asthmopolis inhaler sensor, and many mobile cardiac monitors. Also, there is a strong move (disclaimer, IBM is a big promoter) of the Patient Centered Medical Home (PCMH), which calls for sensors to provide independence to the patient, care to be provided outside the clinic, and constant monitoring for adverse events.
Nike Fuel was all over SXSW, and FitBit is expanding their their product portfolio. IBM is involved in the (young-mom) activity monitor BodyMedia. And here’s an article of the VC behind the gamified (shudder) heart monitor, Basis. And 23andMe’s Linda Avey (@lindaavey) started up a company (with Mitsu Hideishi, @syntheticzero) called Curious, specifically to explore how to make personal health data available to the (non Quantified Self) masses.
And of course, I don’t need to go into the amazing things happening in the Quantified Self world. Here’s an exciting post of a recent meet up. Look at all the projects!
More recently (and what triggered this post) is the news (link via @erigentry) that FourSquare founder, Naveen Selvadurai, is getting into health monitoring. Nothing captures the attention more of investors than some successful entrepreneur entering a new field that most mainstream folks have not thought of.
Innovation from the outside
One thing that hit me is that whatever arises from the fusion of Big Data and personal health measurement, it will come from outside the healthcare industry. Part of that is because in the healthcare industry, they are thinking of hospitals, chronic conditions, FDA, reimbursements, privacy, and so forth. Too much baggage, I think.
I had hoped that payers would see the importance of tracking health outside the clinic, especially for incentives or rebates on paying for care. My thought now is that they will end up buying someone rather than building something.
Which leads me to my realization that folks outside the industry have none of that baggage, have oodles of experience in consumer web services and data visualization, and, perhaps most importantly, view these devices as curious toys that are begging to be played with in interesting and novel ways.
But the challenge is not to alienate the non-techies who are interested in keeping an eye on their health, but are not as driven as the usual QSer. QSers are the bleeding edge of all this. My brainwave here and excitement is how to being this to the rest of us, making the ideas of QS more mainstream.
“It is NOT about the data but how it inspires”
This quote is from Rodrigo Martinez (@rodrigoATCG) from the IDEO healthcare practice. It dovetails nicely with a comment from DJ Patil (@dpatil) that, “OK, so my super intelligent scale tells me I’m fat. But I can see that looking in the mirror.”
“Inspire” is what I’d expect from an IDEO designer, and that’s fine. But for me, it’s really about motivation – how to stay interested in my health, effect the changes that need to be done, and be proud of my accomplishments. I don’t think “gamification” (as in badges and levels) is the answer (and I find it a bit tacky). I think there are more positive ways to motivate people – competition, social pressure, monetary incentives (and of course, I’d love to see the health plans get involved – even perhaps mandating).
What behaviors are we trying to promote by making this data visible to people? And, more importantly for anything that tries to optimize a parameter, what should we measure? Obviously the parameters we measure should represent something we wish to modify (isn’t the whole point about modifying a behavior to maintain or improve health?).
What do you think? Next Monday, I’ll be in Minneapolis participating in a predictive analytics event. I’ll be part of an executive roundtable, a breakout session on predictive analytics, and a general audience talk on the Predictive Power of Big Data Analytics in Healthcare. One thing I will touch on is the fusion of Big Data and personal health monitoring (another reason for this post). I’m sure I’ll get a lot of feedback on this topic – the event host is a device manufacturer with some monitoring devices of their own.
I also wanted to point out that IBM, my employer, published last year an excellent report on “liberating the information seeker” and The Future of Connected Devices (download PDF). I highly suggest you read it if you’re interested in this space.
So, what do you think? Do you use any of these devices, even just a heart rate monitor while running? A pedometer? What do you think will be the way we bridge the intense measuring world of the quantified self to a user of a device who just needs a little help to stay healthy?
Of course, this post is about devices and personal health. I see some interesting opportunities with personal medical records. But that’s a different aspect of Big Data and healthcare and best left for a different post.
I’ll leave you with a video of a well-used healthcare monitoring example at IBM – the Data Baby. Read the case study too (PDF). Disclaimer, I am part of the team that sells the streaming analytics product used for this project.
Image from dpstyles™ (There’s a weird cognection behind this image: I searched for CC images of Nike Fuel on Flickr and came up with this one, which I thought was pretty cool. When I looked at the other images, I realized this was Dennis Crowley’s stream. This post was triggered by a the Selvadurai news item mentioned above. Weird.)
“In conclusion, a time window exists that enables the artificial colonization of GF mice by a single oral dose of caecal content, which may modify the future immune phenotype of the host. Moreover, delayed microbial colonization of the gut causes permanent changes in the immune system.”
Ok. So there’s mounting evidence that rapid colonization of the gut of neonates is important to immune development. Next step is to translate that into real medical therapies. There are immune and gut diseases that afflict newborns (also, in many cases, newborns are bombarded by antibiotics at this crucial time) – what have we learned to make them healthier and also ensure that their immune system develops properly?
“Exposure to microbes during early childhood is associated with protection from immune-mediated diseases such as inflammatory bowel disease (IBD) and asthma. Here, we show that, in germ-free (GF) mice, invariant natural killer T (iNKT) cells accumulate in the colonic lamina propria and lung, resulting in increased morbidity in models of IBD and allergic asthma compared to specific pathogen-free (SPF) mice. This was associated with increased intestinal and pulmonary expression of the chemokine ligand CXCL16, which was associated with increased mucosal iNKT cells. Colonization of neonatal—but not adult—GF mice with a conventional microbiota protected the animals from mucosal iNKT accumulation and related pathology. These results indicate that age-sensitive contact with commensal microbes is critical for establishing mucosal iNKT cell tolerance to later environmental exposures.”
There have been a good series of papers and studies into the “hygiene hypothesis” – that exposure to microbes early in life are actually important for the proper evolution of the immune system. This paper is one more example of that – these researchers were able to show what happened to the immune cells in the but of mice that never acquire bacteria, acquire bacteria only as adults, or acquire bacteria as pups.
I sometimes think of the 1850s-1990s as the Pasteurian Age – we were controlling bacteria to create a sterile world based on germ theory, aseptic techniques, public policies, and, of course, antibiotics. Alas, in the past 20 years, we’ve come to the realizations that we’ve reached (to joke a bit here) “peak antibiotics”, and that the only bugs to survive our clean homes and hospitals and antibiotics were Superbugs.
Now, in the past 5-10 years, I feel we are entering a post-Pasteurian Age, where we are gaining a deeper respect for the bacteria and fungi that share our world (and bodies) and that we are slowly thinking of how we can balance the sterile world we want and the microbe-filled world we need.
Great time to be a practical microbiologist, don’t you think?
“The results also showed that (i) consumption of an FMP containing five bacterial strains was not associated with a statistically significant change in the proportional representation of resident community members within and between individuals; (ii) the appearance and disappearance of strains comprising the FMP consortium did not exhibit familial patterns in the fecal microbiota; and (iii) B. animalis subsp. lactis CNCM I-2494 was the most prominent assayed member of the consortium represented in the microbiota during the 7-week period of FMP The results also showed that (i) consumption of an FMP containing five bacterial strains was not associated with a statistically significant change in the proportional representation of resident community members within and between individuals; (ii) the appearance and disappearance of strains comprising the FMP consortium did not exhibit familial patterns in the fecal microbiota; and (iii) B. animalis subsp. lactis CNCM I-2494 was the most prominent assayed member of the consortium represented in the microbiota during the 7-week period of FMP consumption. Analyses of the fecal gene repertoire over the course of the 16 weeks of the experiment indicated that (i) variations in the functional features of the (fecal) microbiome were less than the variations in bacterial species composition; (ii) there was no significant difference in the degree of similarity in representation of KEGG orthology group functions for a given co-twin at each time point compared to the degree of similarity that existed between co-twins, whereas individual and twin pair microbiomes were significantly more similar to one another than those from unrelated individuals; and (iii) there were no statistically significant changes in the representation of these functions when the FMP strain consortium was being consumed.”
This is a seminal paper in probiotic research. I have seen a ton of papers on this subject, but none were as thorough as this one. The one concern I had was that there was no control for the FMP (fermented milk product) matrix (I don’t know what to call it, but the fermented milk without the bacteria). I still think there might be a positive effect on the gut microbiome from that matrix.
But, these folks saw similar effects in humans who ate FMP and the mice who had only the bacteria that were found in the FMP, effectively showing what the effect of just the bacteria have on the microbiome.
Still, I’m curious to settle once and for all if there is any beneficial effect of the FMP matrix. My main thought here is that 1) we know that the lactose digesting bacteria help in the stomach (as seen in folks with lactose intolerance), but 2) only one bacterial species from the FMP really seems to make it all the way through the gut. Perhaps the matrix helps the microbiome or signals the microbiome to do something? In this study, it was suggested that the bacteria alone are activating specific microbiome metabolic pathway.
Fascinating stuff. Will need to dig into it more. And I just saw that there are some videos of the authors. Should be interesting.
“A life-threatening germ that causes diarrhea and spreads easily from doctors’ offices to hospitals and nursing homes has climbed to historic highs nationally, federal disease trackers warned Tuesday, as they pointed to efforts in Massachusetts that have helped slow the rate of infections here.”
More on this nasty bug. It’s now getting headlines.
“Bacteriophage could be an alternative to conventional antibiotic therapy against multidrug-resistant bacteria. However, the emergence of resistant variants after phage treatment limited its therapeutic application. Our data showed that the phage cocktail was more effective in reducing bacterial mutation frequency and in the rescue of murine bacteremia than monophage suggesting that phage cocktail established by SBS method has great therapeutic potential for multidrug-resistant bacteria infection.”
Biologic warfare at the bacterial level. In some developed countries antibiotic misuse has caused the rapid development of antibiotic-resistant bacteria. Bacteriophage therapy has therefore taken a more important role. Except, it has its issues. In this paper, they work to avoid phage resistance with a multi-phage approach.
“Clostridium difficile has emerged rapidly as the leading cause of antibiotic-associated diarrheal disease, with the temporal and geographical appearance of dominant PCR ribotypes. We have undertaken a breadth genotyping study using multilocus sequence typing (MLST) analysis of 385 C. difficile strains from diverse sources by host (human, animal and food), geographical locations (North America, Europe and Australia) and PCR ribotypes. Results identified 18 novel sequence types (STs) and 3 new allele sequences and confirmed the presence of five distinct clonal lineages generally associated with outbreaks of C. difficile infection in humans.”
A broad survey to understand the nature of this pesky and increasingly common pathogen.
This paper is trying to measure and model the effectiveness of multi-drug antimicrobial chemotherapy.
Antibiotics are notorious for losing effectiveness as the target microbe gains resistance to that single antibiotic. Being able to treat microorganism with multiple drugs is sometimes the only way to manage the disease – as in HIV or TB.
But to be able to create better multi-drug cocktails, we’ll need to better model the contributions of each component.
This paper seeks to show how to measure and prove the effectiveness of the component sin a two-drug system. But I am wondering how we’ll do the same for more than two drugs (HIV anti-viral therapy has at least three components).
“This study demonstrates that the milk-feeding type and the HLA-DQ genotype differently influence the bacterial colonization pattern of the newborn intestine during the first 4 months of life and, therefore, could also influence the risk of developing CD in later life. Breast-feeding reduced the genotype-related differences in microbiota composition, which could partly explain the protective role attributed to breast milk in this disorder.”
Interesting study doing two things: 1) showing an effect of genotype on bacterial populations in the gut – and that they are different for those at risk for celiac disease; and, 2) showing a difference between the bacterial populations of breast-fed and formula-fed children, and a possible microbial reason why breast-feeding protects against celiac disease.
Very cool. And scary how it makes such good sense.
I wasn’t so sure how to describe Treato. It’s really interesting – mixing semantic analysis, social media, medical records, and analysis – Treato is really the next level of how consumers research diseases.
