When is data the ‘right data’?

Published: 17 Aug 2023

Using the right data to support robust healthcare services

Mirek Skrypak, previously Associate Director for Quality and Development, HQIP

“As I was gearing myself up to write this article about using data to develop robust healthcare services, I took receipt of a rather unusual delivery: A golden microphone with Bluetooth and a voice synthesiser. It proved momentarily inspirational, though how is that relevant, you might ask? Well, it prompted me to think about change and, in particular, how using the right data can support the right change.

On seeing my delivery (which was a present for someone else, by the way), I was struck by how much microphones have changed. I researched early microphones. Although it still performs the same essential function as the original from 1878, my new microphone differs significantly. This golden one with buttons, Bluetooth and charging cable etc shows many visible signs of change, most of which are in response to modern-day needs. Which led me to think about change in healthcare services.

Practice won’t change if you don’t have the right data

The NHS is about 70 years younger than the microphone. Just like the microphone, you could say that it still performs the same function (to improve outcomes for those needing medical care). It also needs to change in order to respond to current need (not least in response to an increasingly ageing population). But how do we ensure that any change implemented is relevant, targeted where it’s needed most, and meaningful to service users? In response, I will share some learning that I think is invaluable in relation to data-informed service design within healthcare…

First, practice won’t change if you don’t have the right data. Whatever tools or techniques you use – whether process improvement, error reduction, waste trimming, agile scrum meetings etc – all of it needs to be done within context. In addition, you need to be curious; you need to ask questions and have ideas. Assuming you have all that, knowing what and how to measure is crucial. Simple right? Well, no! There are a number of variables, models, systems, frameworks and contexts etc that you also need to think about. However, I suggest that there is one other consideration that that is more important than all others: There should be no data without stories and no stories without data. The percentages or figures in charts, tables, recommendations and p values etc are all people. Think about it in terms of this fictional example… Let’s say that there is a national target of 80% for a particular process measure which evidence suggests will improve outcomes. There are circa 170 NHS sites, and let’s assume that 75% are achieving this 80% target for a particular clinical area and pathway. Then, let’s say that, in this pathway, there are 100,000 people who receive this care per year. That means 60,000 patients will achieve the target. But, what about the other 40,000? Suddenly this is not so positive. I know which side – or cohort – I would like to be on as a patient.

There should be no data without stories and no stories without data

So how can we improve these numbers? Here at HQIP, we help service providers to find the right data, to take an effective approach to measuring, and to use that knowledge to build consensus for change. We strongly advocate the synchronisation of Quality Improvement (QI) with national clinical audit and confidential enquiries to enable improvement in outcomes at a national level. Find below what I hope is a useful summary of the key questions that need to be addressed by healthcare leaders and their teams when doing this…

As a manager or clinician:

  • Have you collected the data in the right way to be able to identify trends?
  • Do you have the right roles reviewing the data? Who owns it?
  • Do you know your audience? What is it they need to know?
  • Are you at risk of decision makers reacting unnecessarily?
  • How are you presenting your data? Have you presented the data to show a true picture over time?
  • Is there variation, and is this normal or the result of an exceptional circumstance?
  • Do you need to consider if it’s actually the system that needs reviewing, or does the target need to change?
  • Do you need to measure for a longer duration?

As a team or service:

  • Do we know how good we are?
  • Do we know where we stand relative to the best?
  • Over time, where are the gaps in our practice that indicate a need for change ie improvement?
  • In our efforts to improve, what’s working (and what isn’t)?
  • Do we know/understand where variation exists in our organisation?
  • Why are we measuring all this and what difference is this actually going to make to the quality of services?

That’s a lot of food for thought. And I don’t suggest, for one moment, that it’s an easy task. But it is necessary. While I’m not sure that we need a golden, ‘Bluetooth-enabled’ (or the equivalent) healthcare service, we do need to instigate change. I’m convinced – and I wish I had my gold microphone to hand to make this point – that the right data is the way forward in ensuring that those changes are relevant, meaningful and, above all, deliver maximum impact. However, I’m going to give a final word of caution to the renowned statistician David Spiegelhalter, who says: “Signals always come with noise: It is trying to separate out the two that makes the subject interesting.”

This article is one of a number of articles written for CORNERSTONE, HQIP’s free publication, designed to support Quality Improvement. This article was featured in the 2023 edition, which features other articles on patient engagement, patient safety, and the importance of health data (among other topics).

For more information, go to www.hqip.org.uk/magazine.