Thanks, Thomas. I am on board with the idea that audits can have unintended consequences (and it’s something that Polly, Alan and I have considered particularly in raising questions about ‘routine measurement for continuous improvement’). I’m guessing you’re also particularly interested in the idea that patient safety is ‘a property of the system’. For me, one of the questions that this study raises is how do we understand ‘the system’ and which aspects of ‘it’ do and don’t get considered in investigations of ‘it’s’ implications for safety. The system as formally documented is shown to be rather different to the system as enacted in practice, and the differences aren’t just due to occasional infidelities or ‘bad’ actors. In part its seems that the system as formally documented is unwieldy/ impracticable/ perceived as in some ways unhelpful in the circumstances.
For me there is also scope to broaden out considerations about what falls under the umbrella of ‘safety’ here – it could be much more than reliable elimination of certain pre-known possibilities for error. Lots of food for thought!

You are right Vikki, loads of food for thought. First routine measurement for continuous improvement (CI) and, I should add, for audit also. A good Quality Management System will call for non-conformity reporting. That is non-conformities in product/Service and/or process. See clause 10.2 in BS ISO 7101:23 Healthcare organisation management – Management systems for quality in healthcare organisations – Requirements. These records, and the records of corrective action taken, provide evidence for improvement and audit, particularly if good data analysis tools are used. First the ‘system’ (WAD) has to be documented and managed so that the collection of this data can be readily identified and recorded.

I’m not sure that I see patient safety as being ‘a property of the system’ directly. What I do see is that the system (WAD) has to be understood by not just the people undertaking the work but by many other stakeholders as well. The best way of promulgating this understanding is by documenting WAD. Desk top reviews can then be undertaken for a variety of features like safety, quality, improvement, etc. prior to, if necessary, observations. Many additional things have to be taken into account when considering system safety, things like human factors, ergonomics, staff welfare, staffing, etc. Some of these can only be addressed through observation and reviewing process/patient records.

Next up – understanding the system.

Understanding the ‘system’ is so important. As a long serving Systems Engineer, amongst other things, I have developed my own way of both understanding and documenting systems. (I can provide a real illustration at our next webinar if there is any interest). In the NHS currently it seems to me that the details provided are limited. We have Trusts’ web sites where Services offered and various other bits of detail are accessible. Quality Accounts are published annually. Internally there are Patient Pathways and SOPs, NICE, NHSE, and GIRFT guidance but WAD which taken together defines the system is not available to anyone except the individuals undertaking the work. For the ‘system’ to run properly it is dependant on the cooperation, knowledge, skills, moral, leadership, etc., of the workforce, including all levels of employees. Patient Pathways are useful but they are not the system, they are just Patient Pathways. More to follow.

Following on with discussing ‘the system’ – The last requirement of NHS Impact, ‘ Embedding into management systems and processes’ may instigate change. In the self-assessment guide they have a piece on ‘What this looks like in practice’. Starting form scratch this is a phenomenal task! The new British Standard (BS ISO 7101:2023) could satisfy this requirement but I’m sure Trusts would rather ‘go it alone). Hilda Campbell has provided a great story on her involvement in a project that had a go at understanding a particular ‘system’. My view is that to understand any system it has to be broken down into it’s individual processes and then understand their interfaces with other processes.