The Daily Scan

Happy New Year!

Well, we all made it, it is the last day of 2008. Whew. In some ways the toughest year imaginable, and yet in others, one of the best. Certainly eventful; may you live in interesting times, indeed! One of the cool things about 2008 was that I blogged it, so I can go back and read all about it :) And one of my new year's resolutions is to keep blogging. Thanks to all of you for reading and for your feedback. I hope you have a wonderful New Year's, and see you next year!

Scope, Completeness, and Accuracy of drug information on Wikipedia

MedScape have published a fascinating paper on the Scope, Completeness, and Accuracy of drug information on Wikipedia (PDF).

Background: With the advent of Web 2.0 technologies, user-edited online resources such as Wikipedida are increasingly tapped for information.  However, there is little research on the quality of health information found in Wikipedia.

Objective: To compare the scope, completeness, and accuracy of drug information in Wikipedia with that of a free, online, traditionally edited database (Medscape Drug Reference [MDR]).

Methods: Wikipedia and MDR were assessed on 8 categories of drug information. Questions were constructed and answers were verified with authoritative resources. Wikipedia and MDR were evaluated according to scope (breadth of coverage) and completeness. Accuracy was tracked by factual errors and errors of omission. Descriptive statistics were used to summarize the components. Fisher's exact test was used to compare scope and paired Student's t-test was used to compare current results in Wikipedia with entries 90 days prior to the current access.

Results: Wikipedia was able to answer significantly fewer drug information questions (40.0%) compared with MDR (82.5%; p < 0.001). Wikipedia performed poorly regarding information on dosing, with a score of 0% versus the MDR score of 90.0%. Answers found in Wikipedia were 76.0% complete, while MDR provided answers that were 95.5% complete; overall, Wikipedia answers were less complete than those in Medscape (p < 0.001). No factual errors were found in Wikipedia, whereas 4 answers in Medscape conflicted with the answer key; errors of omission were higher in Wikipedia (n = 48) than in MDR (n = 14). There was a marked improvement in Wikipedia over time, as current entries were superior to those 90 days prior (p = 0.024).

Conclusions: Wikipedia has a more narrow scope, is less complete, and has more errors of omission than the comparator database. Wikipedia may be a useful point of engagement for consumers, but is not authoritative and should only be a supplemental source of drug information.

This kind of research is really helpful as more and more people turn to Wikipedia and other online sources of information.  For me a key finding was this: "there was a marked improvement in Wikipedia over time...".  This is one of the key values of online resources, they can be continually updated.  As new drugs come on the market and new things are learned about old drugs Wikipedia can be constantly updated to reflect the latest knowledge.

FDA approves new prostate cancer drug

NowPublic reports FDA approves new prostate cancer drug:

Federal regulators on Monday said they have approved the first new drug to treat prostate cancer in four years.

The injectable treatment from privately held Ferring Pharmaceuticals fights the cancer by lowering levels of testosterone, which promotes the growth of tumors in the prostate.

Food and Drug Administration officials said older drugs in the same class can actually increase testosterone production before they begin lowering them. In studies on prostate cancer patients, Ferring showed that its drug, degarelix, does not raise testosterone levels.

According to the FDA, patients treated with degarelix had testosterone levels comparable to those seen after the testes are removed.

Excellent.

Why early detection is the best way to beat cancer

Wired has a great cover story this month: Why early detection is the best way to beat cancer.

Ovarian cancer, like most cancers, is measured in four stages. Stage I is early, when the disease is contained in the ovaries. In stage II, it may be present in the fallopian tubes or elsewhere in the pelvis. By stage III, it has migrated into the abdomen or lymph nodes. And by stage IV, the malignancy has spread, or metastasized, into major organs like the liver or uterus. For ovarian tumors discovered in stage I or II, the survival rate 10 years after diagnosis is reassuringly high—almost 90 percent—because treatment is straightforward. But survival rates drop precipitously as the diagnosis shifts to stage III or IV, when the cancer is well established and spreading. Here, the survival rate falls to 20 percent and then to 10 percent.

The US spends billions of dollars to save these late-stage patients, trying to devise better drugs and chemotherapies that might kill a cancer at its strongest. This cure-driven approach has dominated the research since Richard Nixon declared war on the disease in 1971. But it has yielded meager results: The overall cancer mortality rate in the US has fallen by a scant 8 percent since 1975. (Heart disease deaths, by comparison, have dropped by nearly 60 percent in that period.) We are so consumed by the quest to save the 566,000 that we overlook the far more staggering statistic at the other side of the survival curve: More than a third of all Americans—some 120 million people—will be diagnosed with cancer sometime in their lives. Their illness may be invisible now, but it's out there. And that presents a great, and largely unexamined, opportunity: Find and treat their cancers early and that 566,000 figure will shrink.

The article certainly makes a compelling case.  But to me it seems there doesn't have to be a choice between early detection and treatment, we must do both to defeat cancer.  There is also important research being done on how to prevent cancer, and that holds a great deal of promise also.

---

Happy Holidays everyone!

survey: Despite Cutbacks, Hospitals Are Moving Ahead With Some IT Projects

According to a survey conducted by the College of Healthcare Information Management Executives (CHIME), hospitals and health systems are taking measured steps to cope with the nation’s financial crisis, including delaying capital projects and cutting capital and operating budgets.  Some key findings:

• More than half of the CFOs/VPs Finance (55%) are experiencing slight or significant delays in accessing capital and expect the financial crisis to last another 12 to 24 months.
• CFOs/VPs Finance are delaying or lengthening timeframes for completing new facilities or facility upgrades (74%), deferring IT equipment purchases (57%) and delaying or lengthening timeframes for implementing health IT initiatives (52%).
• CIOs are responding by implementing longer timeframes for application projects (63%) and reducing spending on outsourced IT services (34%).
• One-third (33%) of CFOs/VPs Finance have cut budgets while one in four (26%) have laid off staff and/or instituted a hiring freeze.
• Almost all respondents (94%) have cut IT budgets by extending implementation time for existing projects and delaying or reducing the slate of new projects.

Still it is encouraging that:

Even as they rein in spending, many providers are moving ahead with strategic clinical IT projects, which they see as integral to achieving greater operating efficiencies as well as other organizational goals such as increased patient safety. Investments in electronic health records (EHRs), computerized provider order entry (CPOE), and medication management remain high-priority projects for about half or more of all respondents.

Adoption of digital pathology will likely be slowed somewhat by the overall economic climate, but will continue because of the strategic value and long-term benefits...

Lab Soft News turns three

Congratulations to Bruce Friedman as his Lab Soft News blog turns three today!  Thanks for keeping us all informed about lab information technology; the changes in those three years have been amazing.  (If you haven't visited Lab Soft News, check it out - there is a lot of good information and discussion.)  Cheers and all the best going forward!

McKinsey: Three imperatives for improving US health care

Of possible interest, here's a report from McKinsey: Three imperatives for improving US health care (PDF).  At the highest level they are:

1. Reverse the growing incidence of chronic disease
2. Eliminate economic distortions
3. Simplify administrative complexity

Nontrivial issues!  Anyway it is a good discussion of the problems, even if the solutions are hard to find… 
 

Google is my doctor

Scott Adams (Dilbert) blogs Google is my Doctor:

About a year ago I started using Google Alerts to tell me whenever someone mentioned Dilbert, me, or anything about Spasmodic Dysphonia on the Internet.  About six months ago I got an alert with a link to an obscure medical publication with a report about an even more obscure surgical procedure for fixing spasmodic dysphonia.  I took that information to my doctor, who referred me to an expert at Stanford University, who referred me to an expert surgeon at UCLA.  Long story short, the operation I read about wasn't as promising as the article suggested, but the final surgeon in my travels had his own version of surgery that had a good track record.  I tried it, and now my voice is normal.  I never would have found that path without Google Alerts.

This is not an isolated story. Increasingly people are taking their medical care into their own hands, and using Google and other online tools to find information and make contacts. A major trend in health care.

visiting RSNA

Last week I attended the annual RSNA (Radiological Society of North America) conference in Chicago.  You may know, this is one of the largest conferences anywhere, attended by over 60,000 radiologists and featuring the world’s largest trade show for medical equipment.  It always takes place just after Thanksgiving, filling Chicago’s giant McCormick center.  The “booths” of some vendors like GE, Philips, and Toshiba are like small cities, with two story buildings, cafés, coffee bars, and tons of meeting rooms interspersed with the equipment and software on display.

Any one time you go to this conference is fascinating, but it is especially useful to go each year and compare to previous years; you can get a feel for the evolution of the industry and the latest trends.  And this year, the impact of the worldwide economic situation and how it is affecting medical equipment and software.

I often regard Radiology as a leading indicator of what may happen in Pathology, and as time passes more and more of the leading Radiology imaging companies are taking an interest in Pathology (this year GE formed the Omnyx subsidiary with UPMC, Philips incubated an internal digital pathology startup, and Siemens invested in Bioimagene).  It is also interesting to see the evolution in common technologies like displays, projectors, and storage arrays.

Of possible interest, some notes from my visit:

• The attendance at the show was about the same as last year, 62,000 radiologists, and about 700 companies exhibited.  It was notable that a number of spaces on the show floor were empty (companies that had paid for space but decided not to attend) and that the far ends of both halls in McCormick center were occupied by “Bistro RSNA”, an eating area, instead of by exhibitors.

• • Some companies noticeably scaled back their presence, such as Siemens, and some had a large presence but still less than in previous years, such as GE.
  • Of course most companies had planned their presence at RSNA well before the gravity of the economic turndown became apparent. It will be interesting to see the effect next year.
  • Even companies with a large presence brought less people. For example one 3D visualization company brought 150 people to the show last year, many of them as a sort of reward (and to maximize contact with customers). This year they brought 50.
     

• Various booth personnel who I asked said that floor traffic was about the same as last year, maybe slightly less. To me the exhibit halls felt like a crowded zoo, as always :)

• • My overall impression was that exhibitors have been affected more by the economic situation than the radiologist attendees.
  • The economy was a major source of discussion, but more from a personal standpoint (how people have individually been affected) than from a business standpoint (how their companies have been affected.
     
• The trend toward more software applications continues.  The large “modality” vendors are still there (X-Ray, MRI, CT etc. equipment), but they are being eclipsed in number if not booth size by RIS + PACS vendors, CAD companies, and various other informatics companies such as voice-to-text translators.  It is really hard to believe the industry can support so many companies, [apparently] undifferentiated from each other, at least at the highest level.  Just about every booth contained a three+ monitor setup with a worklist on the left and various images in various stages of manipulation.  The idea that radiology has gone digital is completely accepted. 
   

• There were a number of “Nighthawk-like” companies, providing outsourced Radiology reading services. Either at night, or offshore, or with some kind of subspecialty. With Radiology being digital, it is clear that Radiology reading can be done remotely, and a whole ecosystem is forming around this. You can imagine the same thing happening in Pathology.

• An interesting innovation; RSNA’s attendee badges included RFID tags. Every attendee’s whereabouts in the exhibit hall were continuously tracked, and supposedly each exhibitor will get a report of who visited their booth, where they came from, how long they stayed, and where they went afterward. Quite a privacy intrusion, I’m amazed they were able to get away with this without more fuss.

• • It would be useful if they provided attendees with these reports; I’d love to see who-all I visited :)

  • Another useful nice-to-have, a PDA-compatible floor map. The RSNA catalog of exhibitors is too large to carry around, and the online version on the RSNA website uses a plug-in which doesn’t work on PDAs.

• 3D imaging in general seems to be a growing market area.  Not only are there specialized companies like Vital Images, but virtually every major PACS vendor has some kind of 3D visualization built in to their PACS clients (often as an extra-cost option).  Particularly amazing is the ability to record “paths” through a 3D image which can be played back as a “flight”.  Who knew that Fantastic Voyage would become a reality?

• It is amazing just how many companies are in the PACS business. They each have an angle to differentiate. One which was interesting is Carestream; they feature their SuperPACS which is an inherently multi-site design with a web-based front end. This seems like the direction of the future. I spoke with a product manager there who expressed interest in pathology imaging; she said they had heard of it (but not alas of Aperio) and would probably integrating pathology viewing in the future.

• • Most people take it for granted that when Pathology images are stored as DICOM images, existing PACS viewing software will “just work”. The idea that new capabilities are needed for panning and zooming through such large images just hasn’t taken hold.
     

• There were a lot of foreign companies exhibiting; I can’t say more than last year, but it really struck me this year. Hardware from China, Japan, and Korea, and software from all over Europe. There were a lot of “not for sale in the U.S.” signs, and a lot of “not FDA-cleared” notices. Are these non-U.S. products being exhibited at a U.S. show for non-U.S. consumers? Perhaps the dollar has dropped to the point where this is feasible. And RSNA does provide a single marketplace where everyone gets together.

• As Radiologists now spend their time looking at monitors instead of film, ergonometric workstations for Radiologists have become more popular. They range from comfortable chairs and adjustable-height desks all the way to $10,000 workstations which adjust in every possible direction and tilt, and have built in soft lighting and music. Will Pathologists need or want such workstations? Not clear…

• Speaking of ergonometric considerations, the controls used for Radiography equipment are custom built and impressive (e.g. the joysticks used to guide X-Ray sensors into position). Makes you think about what could be done to create Pathology-specific navigation hardware for viewing digital slides.

• It was notable that there are more and more companies which specialize in software to take voice dictations and automatically convert it into “coded” text for diagnostic reports. For years Radiologists have outsourced voice dictation (even overseas!), but now apparently more are using software to convert voice directly to text. Presumably the same software could be used for Pathology reports, the challenges are similar (e.g. correctly recognizing complex medical terms).

• Radiology imaging requires a lot of storage (not as much as Pathology however) and there were a number of computer hardware vendors exhibiting storage, including EMC. I spent some time in the EMC booth (that’s me at right with a 1.5PB Cellera array; it could hold over 6M digital slides). It isn’t clear why these high-end storage arrays are 10+X the cost of low-end PC storage. The vendors try to make a case for their value-add (often including esoteric software features) but it just isn’t compelling.

• • Other storage vendors present included IBM, Dell, and HP
     

• The situation with “enterprise” storage solutions is reminiscent of the high-end specialized viewing displays from Barco, Planar, etc. They may be “better”, but they’re not that much better; certainly not enough to justify the significantly higher prices. It is noticeable that a few years ago “everyone” had specialized monitors for displaying Radiology images (the central monitor of typical three-monitor displays was often 16-bit grayscale), but now “everyone” just seems to use vanilla PC flat-panels. Part of the reason for this is that LCD panels have more consistent color than the CRTs they’ve replaced, but partly I think people have just realized that the extra quality of the high-end displays is just not necessary.

Finally I have to note, nothing says “Christmas” to me like a walk down Michigan Avenue in December (downtown Chicago’s premier shopping thoroughfare). There’s snow on the ground, it’s freezing, and your breath looks like smoke, but the street is beautifully lighted, there’s a ton of people out, and the stores are all decorated and warm and welcoming. For one day it’s wonderful.

Dark Report: Momentum Continues for Digital Pathology

The  most recent issue of the Dark Report* contains this nice article: Momentum Continues for Digital Pathology (PDF):

FDA clears reading of Aperio's digital progesterone receptor slides on computer monitor.

Last month, digital imaging in pathology gained additional momentum with the latest FDA clearance.  Aperio Technologies, Inc., now has FDA clearance to market its slide scanning system for reading digital progesterone receptor (PR) slides on a computer monitor.  Aperio plans to file an application next year with the FDA for clearance to use its digital pathology imaging system for the reading of digital H&E breast tissue slides on a computer monitor.

As the Dark Report notes, this was a significant clearance, following a similar clearance for reading Her2 slides on a computer monitor:

While Aperio is the first and only company to receive FDA clearance for a system that can be used for making diagnoses from digital slides for clinical use in pathology, these FDA clearances are important to the wider pathology community.  Each clearance establishes a precedent that makes it easier for other companies to submit digital pathology systems to the FDA and obtain clearance to market those systems for clinical applications.

And as noted, these IHC quantification clearances are stepping stones to filing for clearance for reading breast H&E slides digitally, the first true diagnosis performed with digital pathology technology.  I hope to be able to tell you more about that soon :)

* The Dark Report Intelligence Service is the leading source of useful information about the clinical laboratory industry and the profession of pathology.