The role of the improvement analyst

I am often invited to meetings with project management teams to work out which measurements will show service improvement. As an improvement analyst, I remind people that projects are about making a process change that will hopefully lead to an improvement.

Don’t confuse a change with improvement, the role of the improvement analyst is to prove the effects of the change. This is not a straight forward job; it's iterative and often accompanied by discovering several things that we don’t need to measure.

The role of the improvement analyst

I have found that there are two broad components to being an improvement analyst:

1. Firstly - and probably seen as the traditional analyst role - is to look at what data we already collect and see what we can do with it. 
2. The second is less common and builds upon improvement methodology, using such tools as the 7 steps to measurement, Juran’s diagnostic journey, Pareto analysis and other such tools to clearly understand the project and baseline, using Donabedian’s Outcome, process and balancing measures to create balanced score cards.

What to do with data

Then, at some point in the project, some data will be collected and it will turn to the analyst to make some sense of it. I’ve created a 3 tier view of how I view what they do with this data.

• Tier 1: Bad data presented in a misleading way. 

In tier 1, data is used as such: the number of patients that achieved a target (eg the percentage of patients who achieved the 18 week target) are presented in a monthly red, amber and green dashboard.

• Tier 2: Bad data presented in a brilliant way.

Here, data such as the number of patients that achieved a target (eg the percentage of patients who achieved the 4-hour emergency department (ED) target) but presented in a daily or weekly SPC (statistical Process Control) chart allowing an appreciation of variation.

• Tier 3: Brilliant data presented in a brilliant way. 

This would be using the actual time that patients spent in ED at patient level, in an SPC chart.  This will allow the analyst to answer any questions asked via tier 2 data (ie how capable are we of achieving a target) whilst allowing the analyst to listen to the voice of the system, picking up on any signals in the noise (A quote often used by Walter Shewhart).

Future Hospital development sites

As part of the Future Hospital Programme (FHP), I work closely with eight busy, multidisciplinary Future Hospital development site teams to give them the skills to collect exemplary data and present it in the best possible way. The FHP team shares the latest news from these sites through the monthly Partners Network newsletter and I would encourage anyone looking to find out more about the Programme's improvement work to sign up. It’s clear to me that we have too much tier 1 analysis, and not enough tier 3, in the NHS and through the FHP we hope to buck the trend.