So, one of the reasons things have been so quiet around here is that I’ve been trying to finish my MPH. I’m pleased to say that this week, I delivered my capstone presentation, thus fulfilling my final requirement.
My talk was on a topic near and dear to my heart: What public health can do to better meet the rising tide of antivaccination opinion. I recorded my talk and most of the Q&A, and thought I would post it here for anyone interested.
The short version? I argue that instead of viewing the antivaccination community as monolithic, public health and healthcare institutions should instead take a spectrum approach, segmenting this community into appropriate categories that better lend themselves to engagement and intervention.
I’ll also post my references and further reading list for anyone interested. Enjoy!
Boing Boing has a great write-up on EpiSurveyor, a fascinating open source epidemiological data collections software that works on mobile phones and PDAs, and uses SMS to send data to a central database, which as you might imagine, just hits so many of my buttons at once. Very nifty.
But I will take issue with one thing: The software seems to be designed with the developing world in mind, and the article mostly references examples falling into that category. But it would be a mistake to overlook applications right here in the U.S. and other industrialized nations. For example, after Hurricane Katrina, Bob Joyce and Balaji Ramadoss, two of my colleagues at other preparedness centers, collaborated to rapidly develop a PDA-based health survey tool to use at the Astrodome in Houston. Volunteers took these surveys to collect health surveillance data, and then brought them back to a central office every evening to download the data. A great idea, and a tool like EpiSurveyor would have made it even easier and faster to deploy.
The map referred to below is also mentioned in this interesting article in Smashing Magazine about modern approaches to data visualization. Some of the techniques described definitely skew more towards novel than useful. In particular, it seems to almost exclusively emphasize complex techniques over simple and elegant displays like Tufte’s sparklines. But I would still say this is a terrific thought piece for anyone interested in finding new or better ways of adding a visual element to a large amount of information.
Also in the New York Times this past Sunday was an article by Gina Kolata about statistics about differing ‘average’ numbers of sexual partners between men and women. As she points out, it is mathematically impossible for men and women to have meaningful gaps in their “average” number of sex partners. And she’s right, sort of. As this reply to Kolata by Slate’s Jordan Ellenberg points out, the problem boils down to math illiteracy and linguistics. First, one of Kolata’s examples actually refers to the median number of sex partners, rather than the mean. Secondly, the average non-academic reader of articles on this subject don’t take the word average to mean, well, “mean.” But instead, they take it to mean “typical.” As a pair, these articles provide a good jumping off point for why allegedly objective statistics can be used to make claims that don’t quite make sense or hold together on closer examination.
Many of you may employ the so-called “five second rule,” which posits that if food falls on the floor, and you pick it up within five seconds, it’s still good. (My mother actually referred to this as the “no one saw that” rule.) Well, Prof. Paul Dawson and his colleagues at Clemson University decided to put some actual rigor to this hypothesis, and tested it out.
The good news is that in food dropped on a salmonella-contaminated surface for less than five seconds does, in fact, absorb a lot less bacteria than food left for longer. The bad news is that even in that short period of time, it may absorb enough pathogen to make you really sick, or even kill you. (And of course, the really bad news is that being semi-safe doesn’t actually make it less gross.)
My evolving presentation on Web 2.0 and public health keeps getting bigger, thanks to new sites like “Who is Sick?”. While I’m not going to try and argue that this self-report-based cluster map has epidemiological validity, I think it’s worth keeping an eye on to see if any real trends emerge. It also drives home the point that I make over and over again in my presentation: grassroots information sources on public health are popping up all over, and if the public health establishment wishes to exert influence, they have to become active participants. (Via BoingBoing.)
I’ve been interested to follow Effect Measure’s self-described “failed experiment” in trying to explain a paper on mathematical modeling of antiviral resistance over sixteen blog posts (up to number ten, as of today). On the whole, it’s an admirable effort. The failure, if there is one, is that the effort seems a little unsure of its audience. In the explanatory post, Revere uses the term “a lay audience.” But as I read the posts, it seems to me that these posts are really geared towards a pubic health/epidemiology audience, albeit one unfamiliar with mathematical modeling.
When all’s said and done, I may try an experiement of my own: trying to summarize the work for a truly lay audience, shaving it down to 500 words or so. (Yes, that’s my idea of fun. Pathetic, I realize.)
I’m sitting in the Orlando airport, waiting for my flight back home to Chicago. I’ve been here attending and presenting at the SALT conference (as blogged about previously). Like most such conferences, the sessions had high points, and not so high points. But I wanted to take the time to post about a few things that impressed me the most.
On the first day, I got to see a presentation by Balaji Ramadoss of the University of South Florida College of Public Health. Balaji and I served together last year on an ASPH/CDC Collaboration Group on learning management systems, and he was there to talk about the prototyping methods USF used when building their own, in-house LMS.
Thursday had a couple of highlights. Bob Boufford and Bonita Bray of the University of Alberta gave a terrific presentation on using low-tech/high-impact games for learning, which also pointed attendees to this useful resource page. Later on, Jeff Kissinger, David Rosowski, and Enrique Barquinero of Florida Community College Jacksonville showed off their “untethered” iPod-based Spanish language course developed for military personnel without consistent online access.
Friday was a half day, but included two of the best presentations. Nora Reynolds, Matt McFarling, and Scott Brewster from the University of North Carolina at Greensboro showed off Econ 201, an exclusively game-based online class. (Besides being impressive in its own right, I think it’s the first online class to get it’s own TV commercial.) And finally, ending the day, Peter Fadde from the Collaboratory for Interactive Learning Research at Southern Illinois University gave a fascinating presentation ostensibly about “Low Physical Fidelity/High Cognitive Fidelity Simulations,” but really offering a more wide ranging overview of how researchers identify the cognitive seeds of “expertise,” and unpack those seeds into something that is, for lack of a better term, drillable. (As a bonus to a former psych nerd like me, his presentation heavily referenced the work of Gary A. Klein.)
All in all, a very worthwhile trip.
Update: Oh well, I guess the free WiFi at MCO isn’t as reliable as I thought. Posted Monday after all.
Hot on the heels of Swivel (about which I blogged previously), comes IBM’s new data visualization project, ManyEyes. Like Swivel, it allows anyone to sign up and start sharing data sets. Unlike Swivel, it features an even richer set of graphing and visualization tools to play around with. It’s just an alpha project right now, but at first glance, I’m impressed. (Via Radar.)
Yesterday, Wired broke a fascinating story about a new antibiotic-resistant bacterium (acinetobacter baumannii) that the medical community sees migrating from the military evacuation chain out of Iraq into the civilian health care system. The specific story is interesting, but the general background on the issues surrounding superbugs would make good reading for anyone interested in these issues.
Two interesting items from the latest Bulletin of the Atomic Scientists: Josh Schollmeyer explains why Ready.gov, well, isn’t, and Lawrence Krauss reviews Sharon Weinburger’s new book about fringe science, using the story of hafnium weapons research as its main example.
(And a note to my friends at the Bulletin, in case any of them come browsing: It seems odd to some of us that at this point in time, when your content is more and more relevant, you seem to offer less and less of it on the website every month. Maybe we’ll see a change in 2007?)
I used to link to Damn Interesting quite regularly, and I’m not sure why I fell out of the habit, particularly when they continue to publish entertaining articles like this one: Undark and the Radium Girls.
Recently I was searching for a Mac equivalent to the CDC’s excellent free epidemiology software, EpiInfo. I didn’t find any. But along the way I discovered three interesting, non-related things:
OpenEpi, an open source epidemiology software project that replicates many of EpiInfo’s features, but lacks its ability to import and process a data set.
Indexed, a fun blog of statistics inspired visual commentary (for lack of a better description).
and Swivel, billed as a YouTube for data.
I’m especially looking forward to playing around with Swivel now that I’ll have a little downtime during the holidays.
Tonight marks the launch of the Journal of Visualized Experiments, an online video-based journal for the biological sciences. The idea isn’t just that authors can show neat (and slightly gross) stuff; it’s that video presentation will make it easier for subsequent researchers to replicate and expand on experimental techniques. I suspect Mac users have a considerable advantage.
Nature has more extensive coverage.
I don’t remember whether or not I blogged about Target’s pill bottle redesign a couple years back. If I did, I may well have used the word “smart” to describe it. And it was, in a way, but not like this…
Among the many interesting things I saw at the APHA conference earlier this month was SimPill, a product/service that employs truly “smart” pill bottles to monitor medication compliance. Every time the pill bottle is opened, it sends an SMS message to a server that keeps track of each opening. If the bottle isn’t opened within a specified amount of time, the server sends SMS reminders to anyone specified (the patient, a guardian, doctors, etc.).
Now, the system doesn’t actually monitor pill taking, and one could make an argument that unless the bottle were smart enough to tell whether or not a pill had actually been dispensed, it might not be quite useful enough. But it’s a nifty approach to an important medical issue, and a great example of how SMS and similar technologies are transforming health care and other industries through real time monitoring and notification.
I went through a period in college when I was fascinated by neurological language disorders, particularly aphasias. While the disorder in this case isn’t aphasia, I was drawn in by Scott Adams’ recent story about (he theorizes) remapping his own brain to overcome spasmodic dysphonia. (Yes, that Scott Adams.) Color me skeptical, but intrigued, until we have more information.
Through a project, I recently had the opportunity to become a lot more familiar with the work of the Maternal and Child Health (MCH) community, so I was particularly interested to read this fascinating article about the history of assisted childbirth in general, and the rise of the caesarian section in particular. And of course, I’m a sucker for almost any piece that retells the story of medical and women’s pioneer Virginia Apgar.
I’ve flown twice since the ban on liquids and gels went into effect, and I’m flying again today. Like most people, I’m willing to trade some convenience for security, but also like most people, I wonder about the feasibility of someone actually mixing up an explosive with common chemicals powerful enough to take down a plane.
The Register’s Thomas Greene recently wondered the same thing. Their consensus? Using the most likely chemical compounds it’s possible, but highly unlikely, since few plane rides offer the laboratory conditions, facilities, or adequate time to complete the process, particularly without raising a lot of suspicion.
Now, is it possible that they recent plot involved some as yet unidentified threat that would have been easier to carry out? Sure, it’s possible. But it’s also possible I’m checking my shaving cream tonight because DHS doesn’t employ enough chemists.