Applications

 Today BITool is the system supporting the Diabetes Register at the University Specialised Hospital for Active Treatment of Endocrinology ″Acad. Ivan Penchev″ (USHATE), Medical University – Sofia. USHATE was authorised by the Bulgarian Ministry of Health to host an anonymous register of diabetic patients in Bulgaria. This register contains 28 indicators of diabetic patients including age, sex, codes of diagnoses of diabetes and its complications, diabetes duration, risk factors, data about compensation, laboratory results, hospitalisations and prescribed medication. The register is automatically generated from a Repository of more than 112 million pseudonymised reimbursement requests (outpatient records) submitted to the National Health Insurance Fund (NHIF) in 2012-2014 for more than 5 million citizens, including 436,000 diabetic patients. The outpatient records are semi-structured files in XML format; in each file some tags contain free-text fields with important explanations about the patient: “Anamnesis”, “Status”, “Clinical examinations” and “Therapy”. The BITools functionalities for named entity recognition and extraction of clinical data support essentially the outpatient record pseudonymisation as well as the monitoring of significant indicators like glycated hemoglobin (HbA1c) and blood sugar values. The objective of the Register is to improve the healthcare and quality of life of diabetic patients and their families therefore adequate monitoring strategy was needed.

This application, up to the knowledge of the authors, enabled a unique construction of a medical repository using specific language technologies in large scale, at national level, on an archive of existing patient records without burdening the General Practitioners (GPs) and other medical experts with additional data collection. In Bulgaria the information about diabetic patients was split at various institutions and sources, it was not collected and processed by contemporary information technologies, no databases and registers were available. Therefore no analysis, monitoring and adequate health management decisions were possible. On the other hand building a register by an ordinary approach would be too expensive, slow and ineffective.

The pseudoanonymisation of the patient records was performed on primary data in the National Health Insurance Fund; USBALE received only anonymised outpatient data. Tracking numerous visits of specific patients is possible using the pseudonymised ID of the citizen (EGN). In this way targeted monitoring of particular cases can be done by delivering findings and alerts back to NHIF and other health authorities who can communicate some feedback to the GPs.

BITool provides fast overviews of this large register in different dimensions. The figures below show sample visualisations of Register data:

  • the number of diabetic patients in the dimensions age-gender (at certain moment);
  • the number of diabetic patients per regions.
Here BITool operates on the structured information from the NHIF archive: patient pseudonym, region code, age and gender. Further interesting statistics of this kind might concern explorations of diabetic patients per types of diabetes and diabetes complications, per GPs, per types of medication, according to frequency of visits and so on.

 

Number of diabetic patients grouped by age

Number of diabetic patients grouped by regions


Extracting and processing numeric values of important diabetic indicators like the glycated hemoglobin HbA1c, BITool provides explorations of patient status before and after some specific event or moment of time. The figure below shows changes of glycated hemoglobin levels before and after admission of a specific drug (it is presented in the outpatient record by the drug code). The exploration is done for about 400 patients who had HbA1c higher than 7% (considered normal level) in certain period which is the condition to start treatment with this particular drug (the brown area). It is seen that the HbA1c levels significantly decrease (the green area) therefore the therapy implies positive changes after first admission. However, the typical BI output is exploration of tendencies in the development of some processes. The last figure displays the number of patients who had changes in the HbA1c levels within the interval [-5,5] units for certain period of time. The tendency is that for most patients the HbA1c level decreased by 1 unit. 

Exploring reductions of HbA1c levels after first application of a particular incretin based drug

Reduction of HbA1c levels after application of incretin based drugs