The NHS is funded via a ‘payments by results’ system¹ where hospitals are paid for the work they do instead of being allocated a certain amount of money to spend each year. To demonstrate their work, NHS trusts utilise clinical coding for each patient’s interaction with the health service.

The Structure of the NHS in England

This diagram² outlines how the NHS is structured and funded from the national to the local level. At the top, the Department of Health and Social Care sets the overall strategy and receives funding from the Treasury. Since March 2025, NHS England has been brought directly under the Department’s control.³ Beneath this, Integrated Care Systems (ICSs) are responsible for planning and funding care across regions. At the local level, services are delivered by providers such as hospitals, GPs, and community teams.

Understanding Clinical Coding

Clinical coding is the process of translating medical information from a patient’s record into a nationally standardised code using disease or procedure classification systems such as the International Classification of Diseases, most commonly known as the ICD-10. There may be multiple codes generated for each patient for the same attendance at the hospital based on their primary diagnosis, co-morbidities, and any procedures they may have had. For example, a patient with a background of hypertension, diabetes, and hypercholesterolaemia who attended the hospital for elective total knee replacement for osteoarthritis of the knee would have each diagnosis coded separately.

This could become very complicated for a patient who had a long length of stay at the hospital for an emergency that involved multiple procedures, a new diagnosis, and new treatments. Therefore, to make the interpretation of clinical coding easier, every patient’s clinical codes for one interaction with the health service are placed into a Healthcare Resource Group (HRG), which represents a group of diagnoses and interventions that use a similar amount of NHS resources. Currently, there are 2,782 HRG groups, even though this sounds like a lot, it is much less than 70,000 clinical codes for diseases in the ICD-10!

The Doctor’s Role in Accurate Coding

Most NHS coding departments rely on discharge letters as the main source of clinical information for each patient, which is then used to generate codes that determine trust funding. A study⁴ looking at coding accuracy for patients undergoing day case anterior cruciate ligament (ACL) reconstructionfound 47% of patients had incorrect coding due to errors in coding for diagnoses and procedures, leading to an average loss for the NHS trust of £290 per patient.

They also stressed the importance of accurately describing operative procedures. More than half of the patients in the study had meniscectomies as well as ACL reconstruction, but only 38% had this information in the discharge letter. It was very common for a procedure to be described as ‘Right knee ACL reconstruction’ as opposed to the more accurate ‘Right knee ACL reconstruction + partial medial meniscectomy’. This led to the NHS trust not getting paid for this secondary procedure, a loss of £685.

Enhancing Clinical Notes for Better Coding Accuracy

Discharge letters remain the main source of information for clinical coders in most NHS trusts. While they primarily support safe handover and continuity of care, they also play a key role in how activity is recorded and funded.

To ensure that the care delivered is fully recognised in coding, accurate and complete documentation is essential in the following three key areas:

1. Co-morbidities – Always document all relevant co-morbidities. Where possible, use precise terminology. For example, stating “type 2 diabetes with neuropathy” is significantly more helpful than simply writing “diabetes,” as it allows coders to assign the most accurate code.
2. Investigations – Clinical coders are not permitted to interpret investigation results themselves. Instead, a diagnostic statement from the responsible clinician must be documented. This can be achieved by linking the test to the diagnosis it supports — for example: “BMI of 27, indicating obesity,” or “Echocardiogram showed LVEF of 32%, consistent with heart failure.”
3. Diagnoses – Recorded diagnoses are vital. Phrases like “?”, “query,” or “possible” before or after a diagnosis means it cannot be coded. Instead, use terms coders can act upon, such as “presumed,” “probable,” “treated as,” or a definitive diagnosis. Avoid ambiguous wording like “suspected,” “impression,” or “maybe.”

In some NHS trusts, clinical coders also have access to electronic operative notes, which can be incredibly valuable due to their crucial role in accurately capturing procedures for tariff generation. These notes often provide the most precise account of what happened in theatre, so detail does matter. One helpful way to support accurate coding is for senior clinicians to clearly state the full procedure performed, including all relevant components. For example, rather than simply saying “lap chole,” something like “laparoscopic cholecystectomy with intraoperative cholangiogram” gives coders the information they need to reflect the procedure correctly. Including specifics such as approach, laterality, devices used, and whether it was elective or emergency can make a real difference.

Finally, general clinical notes written during a patient’s stay often provide valuable context. They can help clarify details when discharge letters are brief or when multiple conditions are involved. Structured, specific documentation helps ensure coding reflects the full scope of care delivered.

References

1. Department of Health. Delivering the NHS Plan. London: DH; 2002
2. Powell T. The Structure of the NHS in England. London: House of Commons Library; 2023
3. Streeting W. NHS England: Health and Social Care Secretary’s Statement. Parliament; March 2025
4. Razik, A., Venkat-Raman, V. and Haddad, F. (2013) ‘Assessing the Accuracy of Clinical Coding in Orthopaedic Day Surgery Patients’, The Bulletin of the Royal College of Surgeons of England, 95(1), pp. 14–16.

Written by Dr Amar Sidhu (CT1)