Daughter Anna Jones gets her MFA in Critical Studies from the California Institute of the Arts, and shows off her friend John Merritt’s studio.

New transportation

A call to BC/BS from our office manager elicited the information that they could not cancel this message.  She thought she had to keep stuffing paper into the machine until all 753 pages arrived.

What worked? Unplugging the fax phone line for an hour. I guess the wonderful RightFax server got enough ring-no-answer completion codes that it gave up.

In 1988, disco sucked. It still sucks.
In 1988, fax sucked. It still sucks.  Health insurance fax sucks the worst.  They deny receiving them and won’t stop sending them.

A letter to my students « The Berkeley Blog.

Doctor and Patient – Talking to Patients After a Medical Error – NYTimes.com.

Meaningful use objectives: eligible professionals, hospitals | HealthcareIT News.

See also http://www.cms.gov/EHRIncentivePrograms/

Mike Brittain’s post here about image-grinding reminds me of last year’s project my colleagues and I worked on to handle mass quantities of fax images. His problem was making new sizes of JPEG images to sell handcrafts on Etsy.com.

Ours is making spindly-fonted, half-toned, and blurry medical record printouts easier to read when they’re faxed to somebody.  For some reason, critical information like Coumadin dosages are printed in half-toned gray boxes on some care plan summary documents.  If you just fax them, you can’t really tell the difference between 20 mg and 200 mg. Fail. Fail-to-morgue.  No good.  So we scrub out the the halftones, thicken up the spindly letters, and sharpen up the images people get in their fax machines.

Brittain did his work in the Amazon EC2 cloud using GraphicsMagick.  We built ours and still run it, not in the cloud (duh!  protected health information in those faxes) using the multiprocessor version of ImageMagick.  Both are terrific open source projects.

Do people working in Health IT really love HITECH and ARRA 2009 regulations?  Well, some of them are a bit draconian.  It’s possible for the “corporate veil to be pierced” for data breaches.  That is, individual employees are personally liable for breaches as well as their employers.  That makes me hope somebody will change the passwords and not tell me the new ones.

And, the “meaningful use” provisions are a bit murky as yet.

But this Center for American Progress writeup is a good reminder about why we’re doing this stuff.  We’re trying to use evidence and statistics to improve outcomes while saving money.

Personally, I have been buried lately trying to figure out how to do gnarly tech stuff like self-joins on patient episode tables, and large-scale regressions on geometric average length of stay data. It’s helpful to come up out of the ditch to understand why we bother with this geeky stuff.

Self-joins (intricate SQL language incantations) let us figure out hospital readmissions.  GLOS regression let us measure improvements in outcomes in a decent way.

Here’s an amazing piece of statistical analysis by three researchers at the University of New Mexico, Christopher Eppig, Corey L. Fincher, and Randy Thornhill.

By analyzing datasets gathered by others, they have come up with statistical evidence that national averages of intelligence are strongly correlated to rates of parasite infection.

Correlation doesn’t imply causation, which these scientists know. But, if it can be proven that infections and infestations injure cognitive ability, then our work is cut out for us. No longer will people in wealthy countries be able to deceive ourselves into thinking we’re better. We’re just healthier. No longer will light-skinned people be able to deceive ourselves into thinking we’re smarter. We’re just healthier.

And the challenge of justice and righteousness is to spread good health everywhere. This isn’t exactly news. But the research result in this paper helps make it obvious.

Has there ever been a Nobel Prize in Medicine for health-care IT people? Maybe this is it.