OpenERM & Voice Recognition

[voip4africa]

Hello, I'm trying to promote the use of OpenEMR amongst medical facilities and practitioners/physicians around East Africa.

There's one main challenge we're faced with in a society that mostly relies on paper based record keeping - how to make it an unchallenging task to convert all the paper records that have accumulated over the years in terms of patient records into OpenEMR data.

Another point is, while paper records are more reliable than paper-less, in what form is OpenEMR data stored and how easily can it be backed up onto remote online backup facilities.

Thirdly, knowing that medical facilities and practitioners/physicians here have always kept paper records, a majority of them are too slow at typing and now making them come onto the usage of digital records could be an effort. What level of speech-recognition is or can be included in OpenEMR?

[drbowen]

I can speak for my practice. Even in "paperless world" we still do large amounts of scanning of paper documents. These are named appropriately to allow us to more easily upload them into the electronic record of each patient. A typical document name would be the "patient_ID-type_of_document-date_of_service.pdf" We prefer storing these documents as PDFs primarily because it reduces disk space storage. The default resolution during scanning is 200 dots per inch. This means they don't print out very well but we are making progress in terms of more organizations requesting records in electronic format which means the 200 dpi works out in most cases. If your practices do a lot of printing out of paper documents then you might want to stick to the 300 dpi resolution. Old paper records can be scanned and uploaded in a similar fashion.

Depending on the volume of old paper you want to scan this is still a challenging task. For my practice, I had a lot of storage area, so I just left the old paper charts in the storage area. I walked around with my paper chart in hand for about the first 6 months of using OpenEMR and then I realized that the old paper charts were no longer being looked at, they were only a crutch that I didn't really need, and I stopped carrying them around. Every once in a while I have to go back to the paper to find a particular report, and that's the one report that I scan into the new system. I know this idea of simply abandoning the old paper charts causes anxiety for a lot of practitioners but in reality they only go back to review a very limited number of reports. It saved me a huge amount of work. Eventually, after 7 years we go out and have slowly been destroying the old paper records based on the last date of the visit. Pediatric charts we still keep until the age of majority plus 2 years. Thia means the child has grown up and is now 23 years old.

One of the nice things about OpenEMR is that many different formats can be loaded into the document section. As the viewer selects a particular document, the document is downloaded to the work station and depends on a viewer program on the local workstation to ba able to view the document. As an example, x-rays can be stored as jpegs which can be natively displayed in the application. They can also be stored as DICOM images. Then the work station can have a DICOM viewer installed and still be able to view the images.

Paper charts are not secure from fire or natural disasters. Properly maintained electronic charts are actually more secure than paper records. The current system requires a two process back-up. The OpenEMR MySQL command mysqldump is used to dump the entire database. The scanned documents are maintained in an organized directory under "/var/www/html/openemr/sites/default/documents/". Depending on how the hosting company sets this up one has to mirror this directory to one or more other backup servers. The tools are different depending on the underlying opening system. Windows users will use some type of "ghost' program. Linux and Unix users will likely use "rsync" to accomplish this task.

In the office, have a duplicate server set up with a identical version of OpenEMR. At 2 AM I have a cron job that runs the rsync program that synchronized the "documents" directory. My office takes a daily snapshot of the database around 1 AM local time. In the event my main server "blows up" I just have to restore the must recent snap shot to the secondary server. These snapshots of the data base and the documents are also mirrored to a remote location. Remote locations are chosen for their stability both in terms of weather, geologically, and politically. I like to have my backups where there are no typhoons, hurricanes, tornadoes, floods, forest fires or earthquakes. In Africa, one also has to consider political stability. The backup location may need to be in a different country or even on a different continent. The idea is physical security in spite of what may occur in the local environment of the medical practitioner.

Sam Bowen, MD

[drbowen]

Ooops... Forgot Voice recognition.

The best Voice recognition option here is "Dragon Naturally Speaking" with a medical dictionary. Voice recognition usually gets you about 95% fidelity. To get 100% fidelity you still need a human medical editor to read each note and verify the accuracy of the translation. There are commercial systmes that do better than this but most cost more than I want to spend.

There are also companies that still specialize in dictation for physicians. I work with a company that has a dictation system where the practitioner dictates using a digital dictaphone. The wave file is loaded by software to a "dictation server". A dictation manager pushes the wave files out to the transcriptionists. Documents can then be pasted in the "form_dictation" or uploaded as MSWord Documents or as OpenOffice documents.

I learned to type.

Sam Bowen, MD

[voip4africa]

Thank you, Dr. Bowen for the detailed reply. I appreciate the time you have taken to explain things to the detail.

I did an installation of OpenEMR at a physician's clinic and they've really adopted very well to it. They started with just the front-desk being administrator and handling it all, but limited their usage to basic patient records and billing/payments.
They now seem quite excited about it and would like to do more, especially with the physician adding patient encounters.

On this, I'd like to know more on how OpenEMR handles 2 computers sharing data from a single OpenEMR installation. I'll just broaden on this - the front desk would still be administrator and would handle calendar, patient details, billing/payments & checkout (what is the current installation). Then there's the physician who will (on his own separate laptop) use OpenEMR JUST for patient encounters and things like prescriptions. Once the physician completes with a patient, the patient now goes to the front desk to make payment, based on encounters recorded by the physician. Here the front desk would give the patient prescriptions, lab reort requests, etc., based on the physician's encounter recording. So in short, the physician's laptop would just be reading the patient data from the administrator's OpenEMR installation and what encounters the physician records, would again be stored under the patient records in the administrator's OpenEMR installation?

How would one be able to go about with such a scenario?

Thanks.

[drbowen]

Which ever computer has the fastest CPU and the best RAM would be the best to contain the primary data. I have 17-20 concurrent computers accessing data from one instance of OpenEMR without difficulty. The users need to remember not to try to edit the same enounter note / form simultaneously. it is OK to use the same encounter at the same time by different users as long as they are working on two separate forms. If they unintentionally try to edit the same form at the same time, whomever sves their form last wins (the earlier data will be overwritten). One of the important things to remember, especially if they are using a laptop as the server, laptops do not last as long as desktops and backups or the OpenEMR installation is critical.

Sam Bowen

[voip4africa]

Thanks, Dr. Bowen,

On the same note, how have you installed the OpenEMR on the single instance PC (I believe you have only one PC which has OpenEMR installed on it and the rest of the machines accessing it form there)?
Have you installed it in the shared folders or All Users (in Windows)? What about network security on the All Users - wouldn't it be easily "hackable" from the outside sources like the internet?
What about PC's/laptops with different operating systems accessing the Windows based "central" instance of OpenEMR?

Thanks.

[drbowen]

OpenEMR is a web enabled program. (I actually run Ubuntu Linux on my server.) OpenEMR on MS Windows is usually installed using XAMPP. The web root is in

C:\xampp\htdocs\openemr

Once the openemr is in the web document root Apache and MySQLwill need to running as services.

Assuming the IP address of the computer where OpenEMR is installed is:

192.168.1.1

Then the client computers can access OpenEMR through their web browsers by opening:

http://192.168.1.1/openemr/

This should take your users to the login page at:

http://192.168.1.1/openemr/interface/lo ... _frame.php

If the computer running OpenEMR has a firewall, you will need to have port 80 open.

Sam Bowen
http://oemr.org

[voip4africa]

Thanks, Dr. Bowen.

It works well.
I have OpenEMR installed on my laptop which is WinXP based and it's correctly installed and working using XAMPP.

On my network, I have 2 other PC's - both running Ubuntu and I can successfully access and use OpenEMR on them with full access from the WinXP laptop based installation.

Thanks once again and Warmest Regards.

[drbowen]

Remember to set up regular backups of the openemr data and scanned images.

Sam Bowen

[voip4africa]

Remember to set up regular backups of the openemr data and scanned images.

Sam Bowen

Oh sure.
On the same note, i've registered a remote online backup service (10GB to start off - I checked the size of c:\xampp and it says 322MB) and am going to be using rsync so my Physician client get incremental backups - this way he can save on storage space and bandwidth.
My plan is to just backup the c:\xampp folder

In your first reply to my post, you mentioned

The current system requires a two process back-up. The OpenEMR MySQL command mysqldump is used to dump the entire database. The scanned documents are maintained in an organized directory under "/var/www/html/openemr/sites/default/documents/".

I don't intend to do a MySQL dump because I am hoping that these folders are somewhere under c:\xampp (even though I tried to look for them and could not see any var directory)?
You also mention

At 2 AM I have a cron job that runs the rsync program that synchronized the "documents" directory

Again, I am hoping that the documents direcotry you are refering to, is somewhere under the c:\xampp?
The c:\xampp folder I believe holds EVERYTHING to do with OpenEMR including all patient data, openemr data, scanned images, etc.. am I right and it would be enough just to back this one folder (and everything under it) up?

Thanks.