The patient’s case for transitioning from a paper to an electronic patient record (EPR) is easily stated. Patients want the best care possible, and the best care possible simply won’t happen with a paper record. But for a busy clinician simultaneously juggling multiple tasks, it is not an easy trade-off – giving up the convenience of jotting down exactly what they want to say for the tedium of hunting down the nearest match from a pre-defined list of terms structured for a computer.
It is, nevertheless, critical to choose the EPR over paper, and we must find ways to get UK clinicians there with the least possible disruption to their clinical workflow. The downstream effects of paper documentation — or poor electronic documentation for that matter — are unacceptable, ranging from from care that is at best, ineffective, and at worst, outright harmful.
Consider for example, the COVID-19 pandemic. Without a computerized record, we would not even know the scope of the pandemic itself — much less which complications kill or what approaches are effective. We would not have that information because extracting and analyzing information from paper records requires an army of trained informaticists. Even then, the analysis is suspect and thousands would die unnecessarily while we tried to figure out what is going on, or how to treat COVID-19 more effectively, from a paper record.
The case for clinical interface terminology
For years, the solution has been obvious: Mandate the transition to an electronic record, and make sure every point of documentation simultaneously captures both (1) the clinical intent of the provider and (2) the translation of that “clinician-speak” into machine-processable language.
Still, the transition to an effective EPR in the UK has been…ongoing. In the 1980s, the National Health Service (NHS) adopted a system created by Dr. James Read for structuring General Practice clinical terminology. Over the more than 30 years since, the Read codes have been expanded, improved upon, and ultimately replaced by SNOMED Clinical Terms (SNOMED CT), which owes much of its international core content to the CTV3 iteration of the UK’s Read codes. In 2014, NHS’ Framework for Action endorsed adoption of SNOMED CT as the structured terminology for primary care beginning in 2016, with expansion to the entire health system by 2020. Among the NHS’ data entry requirements for SNOMED CT are to make its terms available as “part of the user interface design” of the [EPR] solution.
But what type of user interface should an EPR have? From a clinical perspective, the primary need is an interface between what the clinician wants to document and the structured, machine-processable code required by the EPR. However, the health system requires many more vocabularies than just SNOMED CT codes. For example, all NHS-based “inpatient episodes and day cases that contain diagnoses must be recorded to the mandated version of ICD,” and ICD-10 codes are also required for hospital diagnoses submitted within a Payment by Results (PbR) program. This is not just some arcane requirement of interest only to academic informaticists – fiscal solvency is at stake. The NHS PbR guide entitles its first chapter: From patient notes to pound notes.
The need for multiple, clinically facing, structured vocabularies – and there are many more beyond SNOMED CT and ICD-10 – slows down the transition to an effective EPR without some sort of interface between the clinician and the computer. Such an interface terminology allows the clinician to express their intended meaning and then simultaneously interfaces that clinical intent to the closest code within each of the required structured vocabularies. If a clinician wants to say, “Pneumonia due to COVID-19 virus,” the interface can serve up SNOMED CT 882784691000119100, “Pneumonia caused by severe acute respiratory syndrome coronavirus 2 (disorder),” or where ICD-10 is needed, U07.1 “COVID-19.”
With a robust interface terminology, multiple, critical pieces fall into place. The clinician never needs to know any of the mapping translations that occur behind the scenes. A different EPR at the patient’s next visit gets exactly what it needs for interoperability. A SNOMED CT clinical decision support rule is properly executed. An aggregated database has structured data that can be readily mined for care improvement. The coder does not need to separately extract a chart into ICD. And, the clinician can just get on with delivery of care.
To learn how IMO can help your organization comply with NHS requirements and document with accuracy and ease, click here.
SNOMED and SNOMED CT® are registered trademarks of SNOMED International.
