Smart Phones and Medicine

Teledermoscopy enables rapid transmission of dermoscopic images via e-mail or specific web-application and studies have demonstrated a high, 91%, concordance between face-to-face diagnosis and remote diagnosis of such images. Further to this, telediagnosis of melanocytic skin neoplasms achieved a diagnostic accuracy of 83% versus the conventional histopathologic diagnosis. Mobile teledermoscopy is the combination of such approaches enabling transfer of images captured with cellular phones coupled with a pocket dermatoscope and preliminary studies have demonstrated the feasibility and potential of its use in triage of pigmented lesions. Such applications are of benefit to physicians in enabling easy storage of data for follow-up or referral of images for expert second opinion and may facilitate a “person-centered health system” for patients with numerous moles and pigmented skin lesions who could forward images for evaluation.

Semin Cutan Med Surg. 2009 Sep;28(3):203-5. Mobile teledermoscopy–melanoma diagnosis by one click?

More data just means more data. Someone – or some thing – has to interpret all of the information generated by new technologies. Too bad we are creating a health care system with all the responsiveness of a snail on downers.

Model that we are creating: A federal bureaucrat commissions a study, to be vetted by a centralized board, to be further vetted by a state panel, to be implemented by a local health care provider, and zzzzzzzzzzzzzzzzzzzzzzz….

Model real-life Millenials will someday use for medical diagnostics: “There’s an app for that.”

I’m exaggerating for effect so don’t get too hung up on the potential accuracy of the prediction. You get the point.

Yell at me in the comments or whatever.

8 thoughts on “Smart Phones and Medicine”

  1. Yes. If we’re lucky people will start using the new model despite, or as an alternative to, the bureaucracy. Maybe we’ll import our new diagnostic apps from India or China.

  2. 9% incorrect diagnosis seems awfully high to me.

    (Yes, I know some of the 9% may be diagnoses that were actually incorrect when the doctor was actually in the room with the patient but I imagine it is a relatively small fraction.)

    If there was a way to move those 9% reliably into the “I can’t tell remotely, you need to see me in person” zone rather than actual incorrect diagnoses then I can see this working.

  3. HCR is not, not, not designed to provide Health Care.

    It’s designed to wipe out the small businessmen/middle class who keep their money by filing ‘S’ returns (self employed) and various small businesses.

    (where for instance 70% of new biotech patents come from).

    That’s why they hired ZERO doctors and nurses but did hire 16,000 IRS agents.

    They’re desperate for money, but that’s not the only motive.

    You know that “Flyover” America they so despise? They also loathe and since the Tea Parties fear that America .

    Welcome to the Rust Belt. For an excellent book on the 70’s and how we dug out of it – Read “Econoclasts”. U of Chitown figures prominently in it.

    “Maybe we’ll import our new diagnostic apps from India or China.”


  4. There are actual systems that will improve care tremendously. About 14 years ago, I was on the faculty at UC, Irvine for a couple of years. A friend of mine was chief of surgery and I sent him some of the work I had done at Dartmouth. There we had used the Medicare database to analyze the care of dialysis patients. I had done a lot of dialysis access surgery and had some strong opinions that I wanted to learn to validate, or refute. Dialysis patients make a very useful database because Medicare keeps a 100% record of all claims. For other Medicare patients they keep only a 5% sample.

    Anyway, we analyzed the care of dialysis patients by studying the claims and then doing a regression using comorbidities like emphysema and heart disease. We were studying the access shunts which are the critical feature of dialysis. No shunt, no dialysis. At the time, most access sites used prosthetic material made by Goretex. It was my opinion that native vein was far better. It turned out that there were not enough patients with native vein shunts to analyze but we were able to analyze the duration of function of the prosthetic shunt until it had to be replaced. Most of these last about 18 months before they clot. Then a new shunt usually has to be placed. The old shunt can rarely be revised enough to last. We were interested to see if diabetes or arteriosclerosis or other co-morbidities affect the duration of function of the shunt. What we found was that the only factor that correlated with the duration of function was the patient’s zip code.

    Thinking about it for a moment, zip code is a marker for either the dialysis center or the surgeon. After I left Dartmouth, some other students used my database to study mortality and found that it correlated with the dialysis center. I presented this a few places but it never got any traction. Nobody is too interested in quality improvement if it looks like it will cost more. Obamacare will be even less interested in quality than the present distributed system.

    I spent a lot of time at UCI trying to get the administration interested in better EMR systems, especially with decision systems built in for places like ICU. Once again, there was little interest. Telemedicine has been used a lot in Alaska. There are a few good programs in Minneapolis for cardiac assessment in small towns that link to the big hospitals. I can’t see this being improved or even retained in what I see coming.

  5. Jonathan – yes, that was the point I was trying to make, albeit sort of obliquely. I wonder if people will bypass the system with new technologies if they can, especially if they find the system cumbersome or unresponsive.

    Nicholas – agree, those numbers are not good enough for diagnostic accuracy, but I can see the technology improving and I can also think of other uses, such as patients documenting skin lesions or keeping a personal history of such lesions. Stuff like that.

    Chris – agreed, the exact numbers do matter.

    Elf – good to see you here! Thanks for that reading suggestion, btw.

    Michael Kennedy – medicine is odd in how it conducts itself. That’s how I’ve always felt about it but I also have wondered if it’s just an “oddball” factor on my part (I mean, maybe I’m the oddball for not “getting it.”) No wonder outsiders look at medicine and scratch their heads. To be fair to us doctors, though, the stuff I hear and read about from consultants seems equally odd. Perhaps it’s because people are looking for things that are “scalable” with the current technologies and not thinking ahead?

    – Madhu

Comments are closed.