Transfer to an electronic type of medical record

Good day, dear!

I wondered for a long time why integrators in most of the implemented systems try not to transfer the historical data of the previous software or archives, but start informatization from the moment of implementation.
Particularly worried about this attitude of integrators in the field of health informatization. This topic is devoted to the consideration of the possibilities of transferring paper thick and “lost” medical records into a normalized electronic form available for analysis.

I will try to anticipate the objections that in many industries data archives are still being downloaded and medical is no exception. Partly agree. He himself led the implementation of Gazprom’s system complex on the territory of several entities where third-party products worked, as well as in many other projects, such as registration of registry offices, a sanatorium complex, the scope of regulation of natural monopolies and others.
In each case, an attempt to avoid importing historical data was undertaken in the maximum possible form. The reason is simple:

1. The procedure is very costly because it contains work with a third-party product (which makes it very difficult to predict deadlines) and work on the verification of the correctness of the transferred data, and this, in turn, hides the complexity of matching the logic
2. Financing of such operations, as a rule, does not constitute large sums.

Thus, this procedure is not interesting either to the developers or to the top management of the customer. And the need to import is revealed already in the process of operating the new system, when all financing flows are distributed (once again I want to remind you that I am writing only about those systems in which I myself participated in the development and implementation).

So, for starters, I would like to highlight the importance of converting paper medical records into electronic form. By electronic form, I mean nothing more than the federal cloud of a unified medical information system, for the financing of which our state has allocated several tens of billions of rubles.
')
I'll start with the obvious and understandable advantages:

• Information about the patient's health is stored in one clear place (cloud), and in case of force majeure, full information about the patient will always be available (allergies to medications, past diseases, etc.)
• Updating the history of the medical card throughout the patient's life. The positive point is that when making a decision, the entire medical history is analyzed from the moment of childbirth (or earlier?)

Of all the obvious advantages may not be:

• Storing information about treatment in a normalized form provides great opportunities for research organizations in finding patterns and assessing the consequences of medical decision-making and implementing large-scale healthcare programs.
• Implementation of quality control procedures for medical care (I will write about this next article)
• Downloaded archives of historical data (paper medical records) will allow retrospective analyzes and accurate predictions at the beginning of implementation.

After we are convinced that the normalization of these medical records is a necessary thing, I’ll proceed to the proposals for implementing data import in the harsh Russian realities:

1. The first step will be the transfer of these medical records into electronic form by scanning into an unrecognized format of a simple image. This task is implemented with relatively little effort. We make an approximate calculation of the cost of digitizing data from one health care facility:
   The number of patients is 20,000. The average size of a medical record is 100 sheets. Relatively inexpensive solutions can be high-speed scanners, the speed of which up to 60 sheets per minute. Scanning does not require highly qualified users.
   Total: need to scan 2 million pages.
   Digitization time = number of sheets / scan speed / number of minutes per hour / number of hours per day = 2000000/60/60/8 = 69 man-days.
   If 20 people are engaged in digitizing data (attracting students is a good practice), then one and a half days will be spent on digitizing one large health care facility with 20,000 patients, which is an acceptable time.
2. The second stage of data digitization is the recognition of the text contained in the patient record. Approaches to text recognition may be as follows:
   - An expensive way - text recognition by a person with competence and rights to prepare such documents
   - Simplified method - using the technology reCAPTCHA (Computer and Humans Apart).
3. Normalization of digitized data. This stage can be performed only by doctors with the knowledge and rights to compile such documents. The task is to spread the recognized (or unrecognized) data in the system, linking them to directories and organizing them into a specific structure. The work is certainly very voluminous and its solution will take time. As one of the solutions, the following is possible (presented in the figure):
   
   • Creating an open medical archives recognition service. On the one hand, there are users who have a set of unrecognized data in electronic medical records, on the other hand, medical workers who have the right to verify and validate patient data. Each data recognition operation is confirmed by a digital signature of the doctor. Such a service can be published in the medical Internet communities, professional medical network, or as an independent service.
   • Financing of this operation is possible from several sides:
     - Federal and regional budgets, funds, investors
     - A patient who wants to recognize their medical records out of turn or in the absence of other funding.
     For recognition of a single card, a reward is declared, which is received by a medical professional who performed the work and who signed it to the EDS. Thus, a large distributed network of workers is created, safety, reliability and responsibility for quality are ensured.

A schematic image of all that is written above:


Summing up, I suppose that it is the creation of the service, in which both suppliers and customers will rotate, will allow to quickly enough accomplish the endless today task of digitizing medical archives and attract the attention of ordinary people to the health problem.

Before communication!

PS Article written by Peter Kondaurov . He will be very grateful if someone has an invite for habr, in exchange for which he promises to publish many interesting articles.
Thank you VolCh for the invitation. All plusominus send pkondaurov