If you look at the solid line in the above graph, which is the combined genetic and non-genetic one, you can pick points on that graph and see what the sensitivity and specificity are. Since I don't have the numbers behind the graph, I have to estimate this by eye, but I think it's instructive. For instance, if you look at 60% sensitivity, that point on the graph is roughly at 55% specificity. So that means if you used that cutoff, you'd be identifying 60% of the cancer patients as high risk and 40% of them as low risk. And you'd be identifying 55% of the non-cancer individuals as low risk and 45% of them as high risk. The dotted line marked 'Gail only' would probably be at about 50% specificity at 60% sensitivity, so the combined model is definitely better. At 80% sensitivity you're only getting about 35% specificity for the combined model, however, and even lower for the non-genetic model. You could do similar exercises for different cutoffs, but my main conclusion is that even though there is better risk assessment with a combined score when compared to non-genetic factors alone, these accuracy numbers are relatively unimpressive. If a physician is going to make choices about patient care, I'd imagine they'd want some numbers much better than any of these results, genetic or non-genetic.
Up at 5AM: The 5AM Solutions Blog
If you look at the solid line in the above graph, which is the combined genetic and non-genetic one, you can pick points on that graph and see what the sensitivity and specificity are. Since I don't have the numbers behind the graph, I have to estimate this by eye, but I think it's instructive. For instance, if you look at 60% sensitivity, that point on the graph is roughly at 55% specificity. So that means if you used that cutoff, you'd be identifying 60% of the cancer patients as high risk and 40% of them as low risk. And you'd be identifying 55% of the non-cancer individuals as low risk and 45% of them as high risk. The dotted line marked 'Gail only' would probably be at about 50% specificity at 60% sensitivity, so the combined model is definitely better. At 80% sensitivity you're only getting about 35% specificity for the combined model, however, and even lower for the non-genetic model. You could do similar exercises for different cutoffs, but my main conclusion is that even though there is better risk assessment with a combined score when compared to non-genetic factors alone, these accuracy numbers are relatively unimpressive. If a physician is going to make choices about patient care, I'd imagine they'd want some numbers much better than any of these results, genetic or non-genetic.
While mutliple items went into its release, the major feature was production integration with Microsoft's HealthVault ( MSHV), enabling full bi-directional data exchange. Anyone who is familiar with the challenge of making technology work - say across companies, cultures, federal agencies, large and small companies, etc. - by a specific date - can understand that this was a major feat, with many people to thank for making it happen. To highlight just a few - Dr. Greg Feero did a great job representing the collective teams and the ultimate value to doctors and patients - both at the Interoperability Stage and the Microsoft Installation (see photos).
From the 5AM perspective, an incredibly focused team grew from our core engineers once we got the "go-for-it" in mid-December to march to HIMSS10 and launch MFHP.
The delivery team coordinated and executed across challenges like data model incongruity, legal clarity, work-flow, data accuracy, browser compatibility (going back to IE6), 508-compliance, and server capacity. Working under a super compressed schedule - we needed to shave 7 calendar weeks to support über-short iterations leading to the conference - we got through the four stages of done, no matter the extended work week nor the multiple blizzards of the century in the DC area, by taking our cue from the following SNL clip on the economy. In the days leading up to the conference, we often needed to turn around fixes across all environments in a day (love that continuous integration). Now that we "fixed it", it's time to "do it".
NHIN Testing - In the IHE Interoperability Showcase, Leslie Power discussed the state of affairs for the delivery of the testing infrastructure that supports the Nationwide Health Information Network (NHIN). The NHIN is a set of policies and technical specifications to enable the secure exchange of health information. With participants like the U.S. Department of Veteran's Affairs, the Social Security Administration, and the Department of Defense, as well as statewide regional health information organizations, the NHIN test infrastructure validates a system's conformance to the NHIN specifications, to validate the exchanges. This collective collaboration has resulted in a suite of conformance and interoperability test tools that will be available in the coming weeks. Download Leslie's presentation here and watch this space for more information on the NHIN.
While MFHP and the NHIN program brought us to HIMSS10, there was an incredible representation of the interests of health IT at the event. Seeming larger than HIMSS09 in Chicago, with the surge of interest brought on by the tremendous opportunities presented by the Federal investments in HIT, walking the cavernous floors was at times overwhelming. Particularly interesting was the reality of the large number of women (especially nurses) representing all segments of power (leadership, management, end-users) who will be at the forefront of adoption, standards, and influence over the value of HIT. The collective brain power present and behind the companies representing themselves to each other and the world should be able to overcome the chasm that has kept HIT from reaching the masses. We should be able to move research and development currently locked in the proverbial basement into the mainstream. For 5AM, we will do our part to help moveavailable technology advances into the hands of users - both clinicians and patients - and create meaningful quality improvements, cost reductions and the creation of informed choice for people and their health. This may represent the dawn of understanding that HIT and the buzz that surround it are tools to make our condition better, as recognized by the people who use it now and in the future.
As we closed the show, we were thrilled when Francis Collins, Director of the NIH, clearly enunciated the URL for My Family Health Portrait - https://familyhistory.hhs.gov/ - and articulated that Family History is the cheapest genetic test on the market on NPR's Diane Rehm show. We are optimistic we will bring together the best of today's technologies and lead the information delivery of tomorrow, showcasing our advances during HIMSS11 next February in Orlando.