Structured data vs unstructured data: What's the future of medicines information?

According to a study conducted by BMJ, 1,700 lives are lost every year in England due to medication errors. In 2020, BMJ reported that 237 million “definitely avoidable” medication errors are made every year in England, which in addition costs the NHS £98 million. 

The healthcare industry in England is still highly reliant on manual administrative processes, such as the manual transcription of prescriptions and patients’ healthcare data. It is estimated that over 750,000 transcriptions of medicines still occur per annum, at various stages of the patient’s journey, such as admission, requesting supply of medicines from the dispensary, and completing discharge documentation, to name a few. 

However, technology is transforming the way in which the industry delivers care, and for it to better support healthcare professionals, pharma, and patients in the future, it is fundamental that we continue to enable the digitalisation of healthcare records and medicines information. 

Quick and easy-to-access healthcare information, such as medicines information, patients records and related data, can help save lives; but for the information to be relevant, it is not enough for it to be stored electronically, it also needs to be shareable across platforms, or interoperable. This can be achieved by formatting it into structured data. 

A quick search on the Healthcare Safety Investigation Branch offers a look at the scope of this issue, documenting errors due to the manual transcription of information, and unavailable patient data, including a case in which an error in the transmission of information across healthcare settings during discharge caused the sudden death of a patient. 

What is structured data in healthcare? 

Structured data is data that has been organised into specific fields as part of a common scheme, with each field having a defined purpose, so that it can be searchable, easily accessible, and shareable. 

The HL7 FHIR specification, for example, is a way of representing data (structured) such as medicine information, patient information and related healthcare data. This standard is considered interoperable and supplies a way of exchanging that information using a defined API interface. 

However, having the structure is not enough, you need to also understand its meaning which requires the use of standardised clinical vocabulary supported in the UK by SNOMED CT. 

What is SNOMED CT? 

SNOMED CT is an international standard providing a global common language for health terms, which all NHS healthcare providers in England must now use for capturing clinical terms within electronic patient record systems. 

To understand how standards such as FHIR and SNOMED CT relate to each other, imagine two systems talking to each other, or even clinicians, the HL7 FHIR determines how they communicate, and SNOMED-CT determines what they communicate. Both are needed in an interoperable scenario. 

What is unstructured data in healthcare? 

Unstructured data is data that “cannot be easily organized using pre-defined structures". 

Unstandardised or unstructured data has limited uses and it is one of the key challenges for the processing of healthcare information. It can negatively affect patient safety, increase workload, create redundancies and, ultimately, allow more opportunities for error. 

Unstandardised data can cause issues when sharing Electronic Health Records between hospitals, GP surgeries, and pharmacies, for example, given the use of inconsistent vocabulary for medical related terms, and the diverse sources of information feeding the systems.

The European Medicines Agency (EMA) has said that “developing an electronic format is the most pressing priority of the actions from a public-health perspective”; they are working to improve access to up-to-date product information on medicines, when and where it is needed, through the development of Electronic Product Information (ePI). 

What is Electronic Product Information or ePI? 

It is authorised, statutory product information for medicines (i.e. SmPC, PL and labelling) created using the common EU electronic standard, which the EMA is planning to roll out in the next 3 to 5 years. 

The information is in a semi-structured format, which means that it contains some structured elements (e.g. consistent, fixed headings and controlled vocabularies), and some unstructured elements (e.g. free text and graphics). 

Is Datapharm’s data structured? 

Datapharm holds ‘semi-structured’ medicines information on emc, aligned with the EMA’s ePI format. We have classed the Summary of Product Characteristics (SmPC) and XML Patient Information Leaflet (XPIL) data into consistent, fixed section headings. This ‘semi-structure’ is in a standard QRD template (QRD is the EMA’s format for SmPCs and PILs). 

As the industry continues to advance towards a common structure for healthcare data, Datapharm is uniquely well positioned to support it, as we hold a nearly-fully comprehensive portfolio of licensed medicines for UK and GB, with a growing portfolio for Northern Ireland and Europe. 

How Datapharm supports accessible medicine information for UK HCPs and patients 

Datapharm’s product information on emc is updated in real-time, upon approval from the regulatory authority. This means that structured data would also be updated in real-time and made instantly available to our established audience of +80 million pharma, HCPs, patients, carers, and other healthcare industry professionals. 

Find out how we can help you make your medicines information more readily available to both internal and external stakeholders, meeting regulatory requirements and in a user-friendly format. 

Get in touch with our team today 

Related content:

[Webinar] The future of electronic product information (ePI) - Watch Now