Saturday, February 18, 2006

Anonymize your data

A website with great healthcare IT resources can be found at the HITSPHERE. This site is maintained by Shahid Shah, a well known Healthcare IT expert, who runs his own blog.

In the current list of information on the HITSPHERE site is a link to a story run by MIT Technology Review. Jeff Jonas, an IBM researcher, reports on ANNA, software that allows for deidentification of individuals stored in a database. Patient information is not just encrypted, but acutally irreversibly changed so that deencryption is not possible. This permits governments and corporate entities to share information in a confidential way.

Does this have an application for public health? As electronic lab reporting (ELR) becomes prevalent in the infrastructure, and the desire for transmission of data increases, institutions may desire more than just reportable conditions.

Examples? Well, NEDSS, Nosocomial surveillance, and the Biosense program for starters. Or sentinel surveillance for emerging conditions using electronic data (i.e. community-associated methicillin-resistant staphylococcal infections, or cancer registries).

But as with most things, it comes down to cost. How much will this cost a public health institution? Can't we imagine a more affordable way? Can't we develop/investigate an open source method?

Sunday, February 12, 2006

Public Health IT in Chicago Works!

Very soon, the City of Chicago will take a big step towards implementing the wired Public Health infrastructure with CHESS, its system of web-based browsers for public health reporting and use of electronic lab reporting (ELR).

Information from the city about CHESS can be found here...

Soon enough, city hospitals will have the choice of using ELR or not, and will train their infection control staff to use web-based forms rather than paper forms for disease reporting.

A recent review explains that ELR has the potential to enhance and streamline reporting, improving completeness. A question that remains is whether the change will be welcome among infection control practitioners (ICPs).

It is likely that the change to web-based/electronic reporting will be accepted by ICP's, because it has to be... workflows for reporting to the city are likely to be permanently changed to use CHESS. However, the degree to which these changes are happily accepted will depend on the effort made to involve stakeholders during development, something that seems to be going well. But for a while, state reporting will be done separately, and it is likely that city reporting will be done on paper and electronically.

This is just another example of the transformative change occurring in public health, where disparate data sources are integrated through good planning and use of IT. Imagine the possibilities: real time disease reporting with case mapping on the web; electronic detection methods to detect clusters of disease (using cluster detection software); prepopulated demographic forms with case information, all sent by labs directly. And goodbye redundant paper forms (eventually).

Saturday, February 11, 2006

An open source public health IT project - HL7

A major endeavor sought by public health deparments is electronic lab reporting (ELR) of reportable conditions. Unlike other measures meant to allow surveillance using hospital data, ELR systems frequently do not require deidentification of patient-level data to conform with HIPAA requirements.

This is because reportable conditions are to be reported between hospitals and public health institutions with patient-level data, and are exempt from HIPAA requirements.

Transmission of data between institutions requires that health care data be transmitted using standard nomenclatures. However, the use of standards, many institutions find, increases initial costs in either time or staff. This is because legacy systems often do not use embedded SNOMED or LOINC nomenclature.

In addition, actual data transmission using ebXML or HL7 methods can also increase costs, as institutions seek off the shelf solutions like Rhapsody.

Seems like a perfect arena for open source software, doesn't it?

One group has such a plan, at least for HL7 messaging: a JAVA library to implement an HL7 reader and parser.

At our institution, we were faced with licensing rhapsody or developing our own solution. We developed our own, using JAVA. Too bad we didn't know about this...

Start of the Public Health IT Weblog

What is this?

It is a new blog to explore and report on issues in the use of information technology in the healthcare industry in the US, with an emphasis on public health informatics.

Do we need another health care IT blog?

I believe we do, because we are witnessing explosive growth in interest and spending in healthcare IT. In addition, the public health IT infrastructure in the US is expanding rapidly. Somebody needs to report on the changes.

Who is the author of this blog?

I am a physician practicing in the Chicago area. After some time in private practice, I became involved in the development of several public health informatics projects.

What will be posted on this blog?

I will post information, links, tips, and updates about projects of interest to those working with IT in public health. I will emphasize sites dealing with implementation of HL7 messaging, standard vocabularies, GIS, and open source solutions. Anyone working in public health knows that costs are as important as quality. We will try to find low-cost, high-quality resources.