Evolution of Medical Informatics

Etienne Saliez, version 2022-06-03 /

• Introduction:
• A review of main problems of healthcare and the contributions of informatics over the years in Europe.
  
• ( 1 )  Statistics:
• Problems:
  
• WHO ICD statistics were difficult to collect and to analyze by hand.
  
• Solutions:
  
• +- 1975: The first medical records focusing on classification of discharge diagnosis. Ministries of health find budgets for these services of public interest. However individual patients do not get any direct improvement.
• ( 2 )  Administration:
• Problems:
• The collection of rendered services and the preparation of invoices is very time consuming.  Moreover many billable items were lost.
  
• Solutions:
• +- 1980: Computer were introduced in most hospitals in order to make invoices. As well in insurance organizations.
  
• ( 3 )  Laboratory data:
• Problems:
• Laboratory results are useful within hour rather than days by traditional postal services.
• Labs are work intensive and more and more tests were required.
  
• Solutions:
• +- 1980: lab results available on-line inside hospitals. 
  
•  +- 1985: begin of automation of laboratories.
  
• ( 4 )  Sharing documents:
• Problems:
  
• Patient need often more than one care provider and communications by means of carbon copies and later photo copies is inefficient, incomplete and time consuming.
• Solutions:
• +- 2005: hospital networks did become common.
  
• +- 2010: regional networks sharing at least some summaries.
  
• ( 5 )  Full paperless patient record:
• Problems:
• Handwritten papers could be scanned but remain difficult to handle.
• Less secretary personnel available due to economic constraints.
  
• Solutions:
• +- 2012: most doctors type their reports directly on the keyboard. Any still incoming papers are immediately scanned.  No more any archives of paper folders.
  
• ( 6 )  Document organization:
• Problems:
• For a long time electronic medical record did remain limited to a kind of simulation of the paper folder but registered in the computer, i.e. visit oriented and sorted on sources of information. Indeed traditionally the only way to manage paper charts.
  
• Solutions:
• +- 2018: Informatics did begin to make possible the navigation in the record from multiple points of views. The possibility to focus on a problem with related information.
  
• ( 7 )  Access to medical knowledge:
• Problems:
• A synthesis of current knowledge can be found in textbooks and in university courses. This knowledge is available as free text.
  
• The conclusion of research projects are converted to guidelines also in free text.
• Most current standards focus on classifications intended for for public health statistics and retrieval of document from the literature.
  
• Solutions:
• 2022: Graph technologies can provide a way to handle complex information. The relations between concepts can be qualified and quantified in a more formal way.
  
• Definition of concepts and relationships can be extended taking account of "clinical modifiers", among others the relative likelihoods.
  
• ( 8 )  Decision support:
• Problems:
• Up to now medical records are still essentially "descriptive" of what did happen. Healthcare professionals are today better informed about the patient data and better access to medical knowledge., However there is still very limited decision support.
  
• The societal problem is that there are not enough qualified doctors particularly in developing regions and in developed regions sub-optimal decisions and even sometimes errors continue to occur every day.
  
• Solutions:
• 2022: experimental solution begin to emerge.
  
• AI as "Artificial intelligence" can provide results as far as very large reference material are available with all the relevant parameters, which is difficult to apply for an individual case. This approach provide results in a blind way based on statistical models.
  
• AI as  "Assisted Intelligence" provide by means of graphs a way to represent the decision process step by step. This can lead to discussion and improvement of  medicalethodology.  This approach provide a way to extend the decision process  achieved by neurons and synapses in human mind. This in a way transparent and easy to understand by means of visual presentation of graphs.
• Of course all previous enhancements remain valuable and must continue.