Section 2: What is Quality Improvement?

Quality Improvement (QI) is a systematic approach to improvement that uses specific methods and techniques to improve quality. The Health Foundation’s publication “Quality Improvement Made Simple” is a helpful introduction and can be found here. Also, see how QI is being used in Wales here.

An essential part of the success and sustainability of QI is the way it is implemented and the approaches used. The key elements to achieving the best outcomes are the combination of ‘change’ (the improvement), the ‘method’ (the approach / the tools) and paying close attention to the ‘context’ and ‘environment’ in which the change is taking place (the people/the place).

There are many types or ‘brands’ of QI to choose from, using a wide range of methodologies and approaches, but many share the following principles to ensure that the ‘change’ is successfully implemented. These include:

• Understanding the problem (and existing data). 
• Understanding the processes, systems and pathways within the service.
• Understanding the demand, capacity & flow of the service.
• Understanding the best approach/tools to bring about ‘change’ e.g., patient/professional participation, clinical engagements, leadership.
• Measurement for improvement, often using statistical process control charts.
• Evaluating the impact of the ‘change’ through qualitative and quantitative measures.
• Understanding the psychology of change and how to lead a change
• Understanding the impact of complexity and the adaptations required to meet cultural and contextual differences.

However, how the implementation of the ‘change’ is managed will depend on the ‘context’ of the service, and this in particular needs careful consideration, and quality checks throughout.

The Institute of Medicine (IOM) suggests that improving quality in healthcare generally involves making it Safe, Effective, Patient-Centred, Timely, Efficient and Equitable.

Table 2 presents the six IOM dimensions and explains why they are considered primary priorities for any NHS intervention/programme and its Research & Evaluation component.

Please note: To ensure that all the six QI dimensions are met, a four-phased research & evaluation approach (discussed in Section 3) would ideally be adopted, using mixed methodologies (discussed in Section 4) and patient and public involvement (PPI) (discussed in Section 5).

There are a wealth of QI technical methodologies, many of which have been used for decades and adapted for use within healthcare. Despite the different names of the QI approaches, most approaches share underlying principles, and many QI methodologies use the same key tools such as the simple Plan Do Study Act (PDSA) cycle described below. Some healthcare organisations choose to use a single systematic QI method, but most NHS organisations tend to choose the ‘best fit’ method for their context. Some of the QI approaches and tools used frequently by TEC Cymru are described below.

Experienced Based Co-Design 

This is a QI approach to improving patient experience of services through patients and professional partnership to design services or pathways.

Data is gathered through surveys, in-depth interviews, observations and groups discussions (e.g., focus groups) and are analysed to identify touch points or themes – aspects of the service that are of significance. A link to the toolkit and useful instruction videos is here.

Model for Improvement (Including PDSA)

This is a QI approach to continuous improvement where changes are tested in small cycles that involves planning, doing, studying, acting (PDSA), before returning to the planning and so on. A link to a how to guide is here.

Each cycle starts with ideas and theories which evolve into knowledge that can inform action and intends to produce positive outcomes. To do this, these cycles are linked with three key questions:

• What are we trying to accomplish?

• How will we know that a change is an improvement?

• What changes can we make that will result in improvement?

Any change that is proposed should also be explained, discussed and communicated with the team.

Statistical Process Control 

Statistical Process Control is a measurement technique that is frequently used to chart data over time. It can help to visualise natural variation (common cause variation) and variation that is not a result of natural variation (special cause variation). The approach uses control charts that display boundaries for acceptable variation in a process.

Data are collected over time to show whether a process is within agreed quality control limits in order to monitor performance and can be used to measure the impact of improvement ideas.

Data & Measurement for Improvement 

Measurement and gathering data are vital in any attempt to improve performance or quality and are essential to assess its impact. It is worth noting, however, that measuring for improvement differs across research & evaluation.

• Measuring for research – tests whether the intervention works
• Measuring for evaluation (or judgement) – helps key stakeholders gauge performance and to collate learning about the process.

When measuring for improvement in terms of QI the learning develops through processes. As a result of a process the key questions or hypothesis will change throughout the project (unlike traditional research). As a result, the data is considered ‘good enough’ rather than ‘perfect’. Instead of asking ‘does it work?’, QI asks, ‘how it works, for whom, under which circumstances and to what extent?’ Ultimately understanding what will constitute success. 

It can be really helpful at the start of any improvement work to map out initial theories about how you will achieve the improvement aim, how you predict change will happen, and what inputs and outputs you expect. There are three useful tools to do this.

1. Driver Diagram: A driver diagram is a simple but effective tool that helps you to translate a high-level improvement aim into a logical set of underpinning goals (drivers) and change ideas. It captures an entire project in a single diagram and also helps to provide a measurement framework for monitoring progress. An example of a driver diagram can be found here.

2. Theory of Change Model: A theory of change is a comprehensive description and illustration of how and why a desired change is expected to happen in a particular context. It makes explicit the underlying assumptions about the project you want to evaluate and provides a visual representation of how your project will lead to the desired impact. It articulates how you expect change to happen and helps to describe the enablers and mechanisms of change. It is also a useful tool to build stakeholder relationships, as you can develop a theory of change collectively using co-production. It can help you communicate your project in a clear and simple way, showing your thinking about what the hoped-for outcome will be. This in turn helps to identify your evaluation and data needs. “Developing a ‘theory of change’ can be useful way of articulating and providing a visual representation of the links between the various activities of service and how this will lead to the long term outcomes it is trying to achieve^” (NPC Guide to Developing Theory of Change) – see here.

3. Logic Model: Logic models describe the relationship between a project’s inputs, activities, outputs, outcomes, and impacts. It can help you to see what you are putting into the project (the inputs), how the project uses the resources (the activities), what products are produced (the outputs), what change is predicted to be achieved as a result of this process (the outcomes) and the final intended and unintended changes that happened as a result of the intervention/programme (the impacts). A useful guide to developing a logic model can be found here.

This traditional QI approach does have limitations however, in that the ‘does it work’ question still needs to be asked e.g. via a Randomised Controlled Trial. It is also important to measure change over time, using methods that make it possible to separate out improvement or deterioration, from the expected level of performance variations.

TEC Cymru split this process is split into four phases across the time period of the intervention/programme. This is discussed in the next section.

To find out more on Quality Improvement approaches and principles see here.