Conway's Law, spaghetti and the NHS

I attended the Clinical Congress Conference 2022 last week hosted by the brilliant Dr Raj Kumar . It was the most successful NHS Clinical Informatics Congress ever with over 500+ attendees in Leeds and via the virtual platform.

I was on two panels and gave the closing keynote on the target architecture for NHS England health and care. One of the questions I was asked in the morning was about the opportunity for transformation with the move to Integrated Care Systems (ICS) and the current merger between NHSD and NHSE. For some reason Conway's Law came to mind, which basically says the architecture, the systems, the products and services you design will inevitably reflect the organisation’s communication structure. The tendency is to design systems whose structure is a copy of the organisations communication structure.??

Whenever I visit the frontline (and I like to do it as much as I can to learn, listen and observe) the effect of Conway's Law on the NHS is plain to see. Patients who cross these organisational and system boundaries are met with a suboptimal service. They must repeat their story because their patient record is not accessible, and the authorised health and care professionals must ask a lot of questions to try and rebuild this picture. Digital shared care records help but there are different levels of maturity out there and whilst some are close nobody yet has a complete record spanning all possible health and care pathways and the patient themselves cannot easily access them. This presents a huge burden in terms of everyone's time, the patient included, and is also prone to errors. I have even seen patients carrying manila folders around with them in anticipation of the questions that will no doubt be asked in fear of missing some bit of information and not recreating a full picture.

NHS England has 42 ICSs that aim to join-up heath and care (primary care, secondary care, community health and tertiary care). Given how these systems have historically been designed, procured and implemented (remember Conway's Law) many are isolated and with this so is the patient information.

So, this brings me onto spaghetti - or more accurately spaghetti architecture.

All of the above creates sprawl of applications, sprawl of data and in the current NHS England landscape there is a lot of sprawl. All leading to the burden of increased time spent on systems trying to do the basics.

?

One way to break down the barriers that Conway's Law has erected is to create smaller multi-disciplinary teams that can work more effectively and deliver better results. However, for these smaller teams to work they need autonomy, be able to innovate, deliver incremental value quickly and establish agile ways of working allowing them to respond to changing needs. I believe national platforms have a role to play here connecting these teams, helping to do the plumbing and giving the teams more time to solve the real problems. To build an effective platform in the NHS it has to bring these teams together with shared goals, a community to share ideas and contribute, an openness to cultivate a sense of "we're all in this together", rather than "it was not built or invented here". There is a huge opportunity to de-duplicate and clean up the sprawl of applications and data silos that add huge complexity, hamper agility and lead to rising costs. Money that would be better spent on our frontline teams.

In my next article I will set the scene in terms of the target architecture, key architecture patterns the key design principles we will focus on to break down these barriers, drive the efficiencies and deliver better outcomes