Data quality: it’s about way more than payment systems

by Paula Monteith

07 September 2021


Let’s face it, the NHS is in the middle of huge change. Change in clinical practice and healthcare provision, driven by necessity. Change in legislation driven by policy. And change in the provider/commissioner landscape. Those of us who have been around forever have seen ‘similar’ before, but not to this extent. So why is data still important?

In what is probably a sacrilegious observation, forget pricing. And for the time being, let’s even put costing to one side (beyond that which is absolutely necessary to keep the wheels going round). But one thing no-one should do is to forget the importance of data, produced in accordance with national standards, and to the highest quality that we can all manage. Because data – what we do, why, and how much of it – is about people… or rather patients.

As we’re all aware, the data in the national system has traditionally been used for pricing and national statistics. But in future, it will be different. Policy makers will use the national data to shape future strategy, believing it to be a true reflection of what’s really going on. Academics will use the data for years, spawning a massive collection of papers, to understand what happened in our NHS in the last year or so, and help plan a roadmap for the future.

And the National Casemix Office (NCO) will use your data as a starting point to improve the healthcare resource group (HRG) classification, to ensure that benchmarking between organisations, within and between providers, commissioners and integrated care systems (ICSs), has clinical meaning.

Without providers adhering to national standards relating to counting, coding and data flow, interpretation of data will be fraught with errors and misunderstandings, possibly leading to suboptimal decisions and patient outcomes. ICSs will make commissioning decisions on something that doesn’t truly represent what is actually happening.

And as we all know, providers are the only people that can make sure this happens, because they’re the ones with the data in the first place.

HRGs are not the answer to everything, (more sacrilege!), but they do have an inbuilt data quality component. The UZ01Z (unclassifiable) HRG effectively provides a measure of how good – or bad – the data you use every day to make decisions is. In essence, the higher the level of unclassifiable activity, the worse the data quality.

Worryingly, in the last year we have seen a decrease in the quality of data submitted by some providers to the national system. To an extent, this is understandable. NHS services have been under extreme pressure and staff have responded heroically in very challenging circumstances. But we have seen patients being recorded without primary diagnoses, age, or gender – some of whom have unfortunately died. How tragic that a person’s final admission to hospital has no details of the why or what they were there for, what the NHS did to help them, or how long they had lived. 

Much of the improvement in data quality over recent decades has been credited to its use in the national tariff system. If a trust wants to be paid correctly, it needs to count activity properly (and consistently) and ensure patient records accurately reflect the patient’s health and diagnoses, as well as any interventions.

So, as we shift from a transactional finance regime to something more ethereal, one final plea – activity and coding information is about much more than payment approaches and funding flows. Good quality data reflects the healthcare we provided to patients who needed it most.

Without standard definitions, and the knowledge that everybody is comparing like with like, we put ourselves into difficult territory. There is a growing recognition that data can underpin the transformation of health and care services, helping us to address the wider determinants of health, and redesign pathways through tools such as population health management, based on patient need and care provision. But this will soon fall apart quite dramatically, if we can’t rely on the data we are basing our decisions upon.


The HFMA has published two briefings recently on data quality

Improving data quality for costing community and mental health services

Costing and data quality: improving the quality of non-financial data required for costing