AI─powered speech recognition technology coupled with the structured clinical terminology SNOMED CT could influence and partake the transformation of clinical coder's role in the NHS─more integrated and part of the clinical team to perform the clinical coding early in the process and help to improve the quality and the consistency of data.
There are many roles in the NHS that we fail to acknowledge, and most are critical to the way the NHS runs. One such role is the clinical coder which is often seen as a backroom function. On my visits to NHS hospitals, I find clinical coding departments tucked away from frontline activity and in some NHS trusts, these teams are even housed in separate buildings.
This attitude to clinical coding is in direct contrast to the significance of the role which is critical to the NHS. It underpins the payment system and is fundamental to benchmarking, allowing hospital trusts to make performance comparisons. If in any doubt, take a look at this Integrated Quality and Performance Report from Guy’s and St Thomas’ NHS FT. It says: “Accurate and complete clinical coding of our activity is important to ensure patient safety, accurate benchmarking and appropriate payment for the services we provide. Improving the quality of all of our data ensures that the information on which we base decisions is reliable.”
This is one of the reasons we decided to run a roundtable for clinical coders in the NHS. This was chaired by Michael Jones, Associate Director of Education (Clinical Coding) Institute of Health Records & Information Management. We also wanted to find out what role technology could play in helping to shape the future role of the clinical coder.
Our roundtable discussion was detailed and gave us plenty of insight into the day-to-day life of clinical coders. We discovered they face many challenges which are neatly summarised in this recent post by Carolyn Cooper, Head of Analytics and Clinical Coding at Guy’s, who also attended the roundtable. One of the most significant challenges arises from coders having to decipher what they find in the clinical records. In many trusts this means looking through handwritten notes.
Where NHS trust have moved over to electronic patient records (EPRs), these systems have matured, and records have become more accessible as a result. However, even where NHS hospital trusts have adopted EPRs, there are still challenges for coders.
For example, it is often impossible to gain a clinical consensus on what is or isn’t relevant and often this means clicking through endless pages of entries in the electronic record surrounded by multiple visual display units.
The consensus from our clinical coding roundtable was that there needs to be a greater understanding of the role and more input from coders when it comes to implementing new systems. Several of our attendees reported that clinician input is sought, but clinical coders are not approached. This adds to the perception that although they have an important role, clinical coders are being ignored.
So, what about the future and the part technology will play in shaping it? Our roundtable coders acknowledge their role will change. This view was supported by our pre-roundtable online survey. Of 135 respondents 87 per cent said the role will change over the next five years. It will see them moving away from being curators to becoming more integrated in the clinical team and involved in predicting care requirements.
Technology will undoubtedly have an impact and speech recognition software could play a part by helping to create meaningful structured clinical terms and pushing these towards coders. One of the catalysts in this respect is SNOMED CT – the structured clinical vocabulary for use in an electronic health record. It is used to standardise the storage, retrieval, exchange and analysis of electronic health data. Secondary care, acute care and mental health trusts have been set a deadline of 1 April 2020 for using SNOMED CT.
SNOMED CT, the world’s most comprehensive medical terminology, is used to standardise the storage, retrieval, exchange and analysis of electronic health data. Representing more than 350,000 clinical concepts, SNOMED CT covers approximately one third of the world’s population through its collection of members and affiliate licensees.
Without use of a structured clinical terminology like SNOMED CT, digital health solutions are challenged to deal with the complexity and variability of unstructured clinical data through the course of care delivery and data analysis. Driven by ontological principles, SNOMED CT’s terminology alleviates artificial intelligence algorithms from having to learn the extent of clinical knowledge from square one, and assists systems to code and access that data in usable and consistently structured formats.
Three quarters of those who responded to our online survey said that SNOMED CT will have an impact on day-to-day activities. Nine out of ten said that technology will have a greater role in helping to improve data quality. Of course, it is perilously difficult to make predictions, but what we do know the future requirements of the workforce will change. The coding profession must be forward thinking and focus on identifying the knowledge that will be needed liaising with clinicians using new technologies, such as SNOMED CT, and upgrading skills in line with changing roles.