Camejo Hospital Medical Records Project
Moderator: peterZ
Camejo Hospital Medical Records Project
Here's an overview of the Camejo Hospital Medical Records Project.
Camejo Hospital is in Leogane, Haiti -- epicenter of the Jan2010 earthquake and the site of Hurricane Tomas landfall yesterday.
Dr Joe and Dr Cam are a married Haitian couple who have been medical doctors in this region for over 20 years. (I know their adult son, Jodel, a software engineer who lives in Colorado with his wife, Sue, who is a nurse.) They have about 80,000 handwritten records of their patients in this region, indexed by a patient ID number. While Dr Joe and Dr Cam are rebuilding the hospital that was destroyed in the earthquake (see renewal4haiti.com for details), operating in a temporary clinic and their currently-flooded family clinic, they are also very interested in converting their patient records into a digital database. We are setting up an open source database for them and considering how to bring their handwritten records into that infrastructure. It's an opportunity to leapfrog to current technology for their medical records, investing in the asset of historical public health information that they have created during their 20+ years of service.
The records are kept in "copy books"; when closed, these resemble large scrapbook/photo albums. I'll post some images here. I'll have the exact dimensions of the copy books in the next few days, but they are approximately 10" x 18" with varying thickness. Blank copy books are made at a local bindery. The index that maps a patient ID number to the patient name is kept in a spiral-bound notebook. Entries are in pen -- black or blue ink -- in cursive handwriting, in French. In the patient exam copy books, there are typically five columns drawn on a blank copy book page; each column is used for one patient during the time that this copy book is active, accommodating multiple visits. Entries are made on both sides of each sheet of paper in the copy book. Entries in the lab copy books are structured somewhat differently than entries in the patient exam copy books. Lab information is transferred from the lab copy book to the patient exam copy book.
The hospital temporary clinic and the family clinic have electricity available sometimes. The earthquake disrupted the local utility grid and it has not been restored. The temporary clinic and the family clinic each use a generator to charge an array of batteries, providing limited AC electricity through an inverter. The temporary clinic will eventually have a solar electricity system that is being designed to meet 100% of its electricity needs; installation will be incremental. Our book scanner must be useable in this environment of limited electricity, without depriving the clinic of scarce electricity needed to provide patient care.
Volunteer and paid labor is available at the hospital temporary clinic to operate the book scanner.
Here are some of the challenges I need to figure out (not all are specific to this DIYBookScanner forum):
(1) How to scan the copy book entries into a PDF digital format, starting with year 2010 and working backward. Need to figure out a suitable camera(s), scanner rig, lighting, power for camera(s) and lights, memory card storage and transfer of data to the U.S. for initial server design and configuration.
(2) How to convert the content of the entries into ASCII text, probably with a combination manual and tool-based process, but probably not OCR because of accuracy concerns.
(3) What open source db tools to use, and how to structure the database, store the info from the copy books, setup patient record-keeping going forward, and train staff.
(4) How to operate the medical record database on a day-to-day basis with unreliable electricity infrastructure and unreliable internet access.
Thanks for any help you can provide.
Sincerely,
DonnaA
Camejo Hospital is in Leogane, Haiti -- epicenter of the Jan2010 earthquake and the site of Hurricane Tomas landfall yesterday.
Dr Joe and Dr Cam are a married Haitian couple who have been medical doctors in this region for over 20 years. (I know their adult son, Jodel, a software engineer who lives in Colorado with his wife, Sue, who is a nurse.) They have about 80,000 handwritten records of their patients in this region, indexed by a patient ID number. While Dr Joe and Dr Cam are rebuilding the hospital that was destroyed in the earthquake (see renewal4haiti.com for details), operating in a temporary clinic and their currently-flooded family clinic, they are also very interested in converting their patient records into a digital database. We are setting up an open source database for them and considering how to bring their handwritten records into that infrastructure. It's an opportunity to leapfrog to current technology for their medical records, investing in the asset of historical public health information that they have created during their 20+ years of service.
The records are kept in "copy books"; when closed, these resemble large scrapbook/photo albums. I'll post some images here. I'll have the exact dimensions of the copy books in the next few days, but they are approximately 10" x 18" with varying thickness. Blank copy books are made at a local bindery. The index that maps a patient ID number to the patient name is kept in a spiral-bound notebook. Entries are in pen -- black or blue ink -- in cursive handwriting, in French. In the patient exam copy books, there are typically five columns drawn on a blank copy book page; each column is used for one patient during the time that this copy book is active, accommodating multiple visits. Entries are made on both sides of each sheet of paper in the copy book. Entries in the lab copy books are structured somewhat differently than entries in the patient exam copy books. Lab information is transferred from the lab copy book to the patient exam copy book.
The hospital temporary clinic and the family clinic have electricity available sometimes. The earthquake disrupted the local utility grid and it has not been restored. The temporary clinic and the family clinic each use a generator to charge an array of batteries, providing limited AC electricity through an inverter. The temporary clinic will eventually have a solar electricity system that is being designed to meet 100% of its electricity needs; installation will be incremental. Our book scanner must be useable in this environment of limited electricity, without depriving the clinic of scarce electricity needed to provide patient care.
Volunteer and paid labor is available at the hospital temporary clinic to operate the book scanner.
Here are some of the challenges I need to figure out (not all are specific to this DIYBookScanner forum):
(1) How to scan the copy book entries into a PDF digital format, starting with year 2010 and working backward. Need to figure out a suitable camera(s), scanner rig, lighting, power for camera(s) and lights, memory card storage and transfer of data to the U.S. for initial server design and configuration.
(2) How to convert the content of the entries into ASCII text, probably with a combination manual and tool-based process, but probably not OCR because of accuracy concerns.
(3) What open source db tools to use, and how to structure the database, store the info from the copy books, setup patient record-keeping going forward, and train staff.
(4) How to operate the medical record database on a day-to-day basis with unreliable electricity infrastructure and unreliable internet access.
Thanks for any help you can provide.
Sincerely,
DonnaA
Last edited by Anonymous on 13 Dec 2010, 23:44, edited 1 time in total.
Re: Camejo Hospital Medical Records Project
Photo of copy books.
- Attachments
-
- copy book stack2.JPG (317.3 KiB) Viewed 20752 times
Re: Camejo Hospital Medical Records Project
Example of index.
Re: Camejo Hospital Medical Records Project
Example of patient record copy book.
Last edited by Anonymous on 23 Nov 2010, 14:00, edited 1 time in total.
- daniel_reetz
- Posts: 2812
- Joined: 03 Jun 2009, 13:56
- E-book readers owned: Used to have a PRS-500
- Number of books owned: 600
- Country: United States
- Contact:
Re: Camejo Hospital Medical Records Project
What an incredible project.
I can see a few challenges that will be tough, one of them being the strange format. Your cameras will be rotated 90 degrees from our standard setup, to capture the "long" pages.
Another is that they've written the records well into the gutter, which means you'll need to use very thin platen glass and platen construction will be important.
The photo of what the books look like -- that's not all of the books you're interested in, is it? Are there more?
I can see a few challenges that will be tough, one of them being the strange format. Your cameras will be rotated 90 degrees from our standard setup, to capture the "long" pages.
Another is that they've written the records well into the gutter, which means you'll need to use very thin platen glass and platen construction will be important.
The photo of what the books look like -- that's not all of the books you're interested in, is it? Are there more?
Re: Camejo Hospital Medical Records Project
Thanks for your observations, Daniel. These photos are not showing all of the copy books. There are many. I'm not even sure how many. My guess is approximately 200 copy books.
By the way, I do have a few very clear, readable photos. The ones that I posted with patient data are blurry because I don't want to post readable data until I redact patient ID information from the photos.
-- DonnaA
By the way, I do have a few very clear, readable photos. The ones that I posted with patient data are blurry because I don't want to post readable data until I redact patient ID information from the photos.
-- DonnaA
Re: Camejo Hospital Medical Records Project
Correction to my earlier post, "The index that maps a patient ID number to the patient name is kept in a spiral-bound notebook."
A more accurate statement is that the indices that map a patient ID number to the patient name are kept in spiral-bound notebooks and also in composition books. The photo I posted is one of the indices that is in a composition book.
Sincerely,
DonnaA
A more accurate statement is that the indices that map a patient ID number to the patient name are kept in spiral-bound notebooks and also in composition books. The photo I posted is one of the indices that is in a composition book.
Sincerely,
DonnaA
- daniel_reetz
- Posts: 2812
- Joined: 03 Jun 2009, 13:56
- E-book readers owned: Used to have a PRS-500
- Number of books owned: 600
- Country: United States
- Contact:
Re: Camejo Hospital Medical Records Project
From my question about cameras, I am thinking that the EP-1 is the camera for the job. Misty took a sample image for me, and I think it looks great:
Misty wrote:Instead of the atlas, I ended up photographing a book quite similar to the copybooks that DonnaA was speaking about. It looks like my biggest challenge is just avoiding camera shake when there's more front-heavy weight than I have with the G10! The slight fuzziness seems to be an artifact of that, particularly as it got worse the longer I kept the shutter speed up.
Here is a full-size JPEG.
Re: Camejo Hospital Medical Records Project
Thanks, Misty and Daniel. I greatly appreciate your expertise in helping to select the best camera for this task. I will check out the camera that you recommended.
Also, the dimensions of a closed copy book are: 44cm x 29cm .
-- DonnaA
Also, the dimensions of a closed copy book are: 44cm x 29cm .
-- DonnaA
Re: Camejo Hospital Medical Records Project
The sample photos I took were of a book which is similar in proportion to your copybooks, just a bit physically larger - it was 59cm x 37cm.
I'm glad to help. Please let me know if I can be of assistance in any other way.
I'm glad to help. Please let me know if I can be of assistance in any other way.
The opinions expressed in this post are my own and do not necessarily represent those of the Canadian Museum for Human Rights.