gms | German Medical Science

Introduction: Hospital information systems (HIS) like the MEONA HIS are commonly used in German hospitals to document medical information. MEONA uses medical catalogues, such as the ABDA drug database, which provides detailed information about drugs, to store all patients’ medication information [1]. Healthcare professionals can prescribe and administer the drugs available in the catalogue and specify the dosage, which describes the frequency and dose of a medication, e.g. one tablet daily at 8 o'clock in the morning [2].

Since 2018, the MEONA HIS has been in full use at the University Hospital of Giessen. For this work, we transformed the documented MEONA drug dosages into the interoperable standard Fast Healthcare Interoperability Resources (FHIR), which makes this data accessible in a standardised way, e.g. for research projects.

State of the art: In Germany, drug dosage information is usually provided using a default pattern such as '1-0-0-0', which represents the amount to be taken in the morning, at noon, in the evening and at bedtime. This pattern can be found, for example, in the Federal Standardised Medication Plan (BMP) [3]. Additional dosing information or information that deviates from this pattern is usually provided in text form and is therefore often not documented in a structured way. Nevertheless, the MEONA HIS stores all complex dosage information in a structured way in its database [2].

The FHIR resource MedicationStatement represents the prescription of a drug and includes information about the scheduled dosage. In addition to the default pattern, all kinds of different schedules can be represented (e.g., intake five times a day or intake only when needed) [4].

Concept: The first step was to analyse what information is required to display drug dosages in FHIR. To do this, we examined the MedicationStatement resource of the Medical Informatics Initiative core data set module 'Medication' [5].

After reviewing the necessary data, we analysed the MEONA database, identified the attributes that affect the drug dosage and transformed them into the FHIR format. The determining attributes are the prescription rhythm, the dosage pattern and whether the drug should be taken before, with, after eating a meal, or on an empty stomach.

Implementation: For the transformation, we used the Extract, Transform and Load (ETL)-Software Mirth Connect in version 3.9. We extracted the available data from the MEONA SAP MaxDB and transformed the respective information into the FHIR field ‘dosage’ of the resource MedicationStatement. We then sent the resources to a FHIR-compliant server.

Lessons learned: Before starting this work, we assumed that almost all dosage information in the MEONA database would be provided using the default pattern, as implemented in the BMP. However, we found that in our HIS only 56% of all prescriptions follow this pattern. The remaining 44% follow other patterns, the most common being 'at any number of daytimes' (15%), 'only at one specific daytime' (9%) or 'take as needed' (9%). This information is particularly important when analysing standardised medication data, as the data analysis needs to be consistent with the data format and ontologies.

The authors declare that they have no competing interests.

The authors declare that an ethics committee vote is not required.