Improving the recording of problems and diagnoses in the NHS

Dr Anoop Shah is a THIS Institute postdoctoral fellow and clinical lecturer at University College London, and honorary consultant in clinical pharmacology and general medicine at University College London Hospitals (UCLH). His research focuses on improving the recording of problems and diagnoses in electronic health records.

As a clinician and researcher, I have experience both of entering data into clinical systems and trying to use the data entered for research. The difficulties I encountered in both roles have spurred me on to try to influence the way systems are designed, so that we can provide better care for patients now and do better research to improve future care. In the last year of my registrar training I was a clinical fellow of the Royal College of Physicians Health Informatics Unit, and I had the opportunity to lead on a project with the Professional Record Standards Body (PRSB) to develop professional guidance for recording diagnoses.

The problem of defining a ‘problem’

The project brought together clinicians and informaticians from a wide range of specialties, and generated much debate even on seemingly straightforward issues such as the naming and definition of key concepts. One of the main problems was the word ‘problem’ itself, which has many meanings, and understanding how it is similar or different from a ‘diagnosis’.

Improving system design and influencing clinician behaviour – which comes first?

Current electronic systems have differing levels of functionality, which affects how clinicians are able to use them. In creating guidance for clinicians, what level of functionality of the system should we expect? How can we write guidance that is realistic and usable today, but also creates the impetus for systems to improve? The process has to be driven by clinicians, as previously information models have been specified by vendors and unstandardised.

Information sharing and different versions of the ‘truth’

Clinical understanding of a patient’s condition evolves over time, and different clinicians may have different impressions. How can these be documented clearly in a shared problem list, and who is responsible for ensuring the list remains accurate, relevant and uncluttered? What do patients think about shared problem lists? I had the opportunity to work with the UCLH Biomedical Research Centre Patient and Public Involvement Group to explore these questions. I found that patients were able to understand the issues and were keen for a solution that enables data sharing with clear provenance.

Next steps – setting up a Special Interest Group

This is an exciting time in health informatics, with interoperability high on the agenda, the establishment of NHSX and shared care records being developed in many parts of the country. However it needs clinical leadership and careful consideration of the way that problems and diagnoses can be recorded and shared. The final report from the PRSB project will be published soon, but it is just the first step along a journey to better recording, which will require sustained effort with multiple stakeholders on board. The Faculty is establishing a Special Interest Group to explore these issues further – please get in touch if you are interested by emailing