Investigation methodologies - please help colleagues in Australia

Dear hub members

We've a request to help from New South Wales.  They and their RLDatix colleagues request:

The public healthcare system in New South Wales (NSW), Australia is changing how we investigate health care incidents. We are aiming to add to our armoury of investigation methods for serious clinical incidents and would love to hear your suggestions.  Like many health care settings worldwide, in NSW we have solely used Root Cause Analysis (RCA) for over 15 years. We are looking for alternate investigation methods to complement RCA. So we are putting the call out  …

Are there other serious incident investigation methods (other than RCAs) you would recommend? What’s been your experience with introducing and/or using these methods? Do you have learnings, data or resources that you could share? Do you have policy or procedure documents about specific methods? Any journal articles – health care or otherwise – that are must-reads?

We've many resources on investigations on the hub and recent thinking in the UK and internationally that might be of value including:

• UK Parliamentary report - Investigating clinical incidents in the NHS and from that the creation of
• A Healthcare Safety Investigation Branch applying a wide range of methodologies in national learning investigations informed by ergonomics and human factors 
• UK's NHS Improvement recent engagement on a new Serious Incident Framework (due to piloted in early 2020)
• Dr Helen Higham work with the AHSN team in Oxford to improve the quality of incident investigations
• Patient engagement in investigations 
• Lessons to be learned from Inquiries into unsafe care and reflections on the quality of investigations
• Insights by leading investigators and resources written specifically for us by inclusion our Expert Topic Lead @MartinL

Do check these out in this section of the hub https://www.pslhub.org/learn/investigations-risk-management-and-legal-issues/

Please add to this knowledge and give us your reflections. We'd be happy to start up specific discussions on topics of interest.

Thank you all, Helen

Hi all, do you think you could help with this. You all have investigation of some form or another as part of your job.

@Gethin, @Bryony Cairnes, @Keith Bates, @Gill T, @Helen Jones, @ZMunson, @mattmansbridge, @Amanda, @Helen W

All contributions most gratefully received!!

Thanks, Clive

Being the hub HF and investigations topic lead I would say… “What’s not working, and why do you suspect that to be the case?”

I would take three or four investigations from last year and see if the interventions worked, or if the report was read, or if anyone still smiles at you!  Revisit the four incidents and see if the system fails to safe, equipment is usable, and patients are now safe. Remember the only reason to do an investigation is to stop it occurring again. If the report is only used to keep a door open and nothing has changed – well let’s do something different.

You may have thought – we do Root Cause Analysis (RCA) and after a good talking to ourselves we realise its nonsense. Indeed, you are correct. RCA is only now found in medicine and it’s what a management consultant tells professionals to do. 

I keep referring in my blogs to Prof Wiki. Breaking my rule to undergraduates that it’s to be treated with caution, to postgraduates that its not to be used, and to post-docs well you should be cast out and your slide rule broken over someone’s knee. But when it comes to RCA the Prof is correct there are lots of problems with it.

RCA never delivers solutions and its pretty pointless. The idea of a graph with more dimensions than a science fiction novel is not good. I’ve done a bit in medicine and I was asked what I thought – just before a person went on a course. I enquired after what she thought of her week long indoctrination. After short pause she described it as like ‘Postcoital depression’. Although it appeared to be a good idea, it did not deliver, had no future use or potential, and a single method and outcome is just mad. As an engineer, I did not know about such medical conditions, indeed none of my partners did – honestly, no complaints in writing, but Prof Wiki description of that does sound like RCA.

When you look at why interventions did not work, think about who investigated, where the facts came from and why you did it in the first place.

Perhaps share with others on the forum an investigation and if it did or did not deliver.

Martin is right. We have created a new 'investigative' industry in healthcare with limited/poor outcomes/data.  The message from many healthcare organisations is that the 'same things happen again'.  That suggests that we haven't understood the underlying problem(s), let alone fixed it/them.

Chris Brougham & I would be happy to discuss our experience building some new technology to support incident investigation.

I have currently designed an investigation method and a training model which is under consideration with the trust a step/method approach, as soon as it has been agreed I will happily post.  It has moved away from RCA per se, but still retaining it as an analytical tool within alongside other approaches.  The method looks at investigative foundations, investigative methodology, witness interviewing, support tools and methods of analysis, hierarchy of learning and recomendations, supporting families, staff and establishing what is need for systems change.  I am trying to make it a more flexible investigative approach that can respond to incident at hand.

As aluded to above I agree from my time here there are repeated themes/events/learning and beyond this method I want to look at systems contsraints that enforce the necessary behaviour changes e.g. People, Systems and Tasks

Keith

Investigations & Learning Specialist

Royal Wolverhampton Trust 

As just a conceptual observer of RCAs, these reads by US authors immediately came to mind when I saw this thread. These authors have tried to examine the RCA process or build out the model to make it more effective. I will add the resources to the hub area referred to above but list them here now due to keep them close at hand for the conversation:

RCA2: Improving Root Cause Analyses and Actions to Prevent Harm. Boston, MA: National Patient Safety Foundation; 2015. 

Hagley G, Mills PD, Watts BV, Wu AW. Review of alternatives to root cause analysis: developing a robust system for incident report analysis. BMJ Open Qual. 2019 Aug 1;8(3):e000646. This review is likely to be on point as it lists tools identified by a literature review that sought to highlight RCA incident review alternatives to RCAs. 

Two PSNet articles that provide background :

Root Cause Analysis Gone Wrong: 2018

Rethinking Root Cause Analysis: 2016

I hope these are helpful in feeding the "fire"!

Lorri

Fabulous feedback, thanks. 

@lzipperer I think the PSNet articles are great, especially the RCA gone wrong. There seems to be much criticism of RCA as a tool that doesn't reflect the organisations and cultural context within which its applied. 

@Keith Bates Looking forward to hearing more of the investigation and training model. We'll be delighted to post to the hub when you're ready 

@Ed Marsden 'Chris Brougham & I would be happy to discuss our experience building some new technology to support incident investigation.' Yes please! Can you write a blog for us? The thinking behind your eva investigation tool, I'm thinking

 

Hi all

I don't think the documents and the detail will be ready until next year buut I have modelled out an approach taken from my own background and researching a number of sources and unapoligetic for plagerising some ideas also.

This short pdf is about my designing of a process that can be adaptive to change and develop some consistency across our trust in due course.  I would of course appreciate alternative views, constructive comments and ideas, as I say it lacks areas of detail which I can elaborate on if required as I say just an outline.

 

KeithPatient Safety Investigations Model design.pdf

 

Hi, Sorry meant to attach this scoping document have converted to pdf and hope it is of some use.  RCA in this method is just an analytical tool and we are moving away from the term root cause Overview Patient Safety Investigations.pdfOverview Patient Safety Investigations.pdf

Thanks for sharing this @Keith Bates
Do you think that along with some commentary this ought to go on the Learn section too?

This recent report from New Zealand looks at restorative approaches and may also be of interest:

A restorative justice innovation: Responding to harm from surgical mesh in New Zealand (December 2019)

Thank you for sharing it on the hub @Jo Wailling

Thanks PSL team.  A restorative approach considers safety systems as made of people and relationships as well as rules and processes. You can check out a brief summary here that is specific to mental health settings. https://www.hqsc.govt.nz/blog/using-a-restorative-approach-to-respond-to-adverse-events/

Learning reviews are also much better than RCA as they focus on storytelling and appreciate context (assuming there is  a safe space to speak up and be heard. )

Thanks Jo. Great blog. I've added it to the hub library: 

 

Thanks this is interesting having worked with restorative justice practioners I am (subject to budgets!) looking to adopt this into the learning environment so a useful debate subject

 

Keith