Now, Video

Shifting the paradigm from technology for organisations to technology for people – Dr Penny Kechagioglou, Lee Rickles and Paul Deffley on patient flow and virtual care

For our panel on virtual care and patient flow, we welcomed Dr Penny Kechagioglou (chief clinical information officer and deputy chief medical officer at University Hospitals Coventry and Warwickshire); Lee Rickles (chief information officer at Humber Teaching NHS Foundation Trust); and Paul Deffley (UK chief medical officer for Alcidion).

The panel covered virtual care and patient flow from an NHS organisation, ICS system-wide and supplier perspective, with the opportunity for our panellists to discuss projects, challenges and approaches along with the technical aspects to consider when implementing digital care. In addition, the panel explored what good looks like for patient flow and virtual care for a digital ICS, and key opportunities for this technology to support integrated care.

Introductions and background

Penny: I’m a clinician by background and a consultant oncologist, so I’m really embedded within the clinical practice and understand what the challenges are within the organisation and the wider ICS. At University Hospitals Coventry and Warwickshire I’m currently leading the electronic patient record implementation within the trust, working very closely with other acute providers within the ICS as we implement our joined EPR. Patient flow is a very common issue which affects all the organisations and it’s important that we work together, integrating our services and making good decisions together regarding how we tackle that issue.

We need to think about why patient flow is important. We know that our population is getting older and we know that up to 50 percent of the population suffers from at least one-term condition. Ultimately, those people need hospital admissions if we are not proactively managing their conditions, but as we know there are a limited number of beds available in hospitals. From local studies, we know that one third of patients do not need medical admission; their needs could be managed at home.

During the pandemic, we used virtual clinics, remote appointments and patient monitoring; we used a range of technologies and worked with partners to keep patients away from acute hospital to minimise infection. So we had a problem, and now we also have learnings available. Now, it’s a question of how we pull those two factors together with the governance of the ICSs. How do we make the most of the ICS structure to embed innovation and improve patient flow?

Lee: At Humber Teaching NHS Foundation Trust we are currently in a process with our wider system colleagues for replacement of our electronic patient record, moving to a second-generation deployment. We used to be one of Digital Aspirant Plus innovator sites until everything got rebranded.

I’m also the director for Interweave and programme director for the Yorkshire Humber Shared Care Record; I cover care records across six ICSs. On that note, we’re slightly different to most other shared care records as we built our own product set with SMEs and vendors, using a lot of open source, platform and standards. This means that we can work in a really agile way.

Paul: I’m UK chief medical officer for Alcidion and a clinician by background, I worked in the NHS for 20 years and I continue to practise as a doctor. In terms of experience, I’ve led digital transformation and programmes across a number of organisations including NHS providers, not-for-profits, social enterprises and national organisations. Prior to joining Alcidion I worked as medical director in the clinical commissioning group structure and was chief clinical information officer for a group of CCGs.

Alcidion is an Australian organisation that has existed for about 20 years with an exclusive focus on health technology. We work with around 20 percent of NHS trusts, ranging from individual modules all the way to flow systems and full EPRs.

Patient flow and virtual care in Coventry and Warwickshire

Penny: As an organisation, working as part of the system, we have been growing our experience of using digital technology – we’re providing a virtual ward service called Hospital at Home, for example, which has expanded to cover patients with COPD, frailty and heart failure. Our experience has been that it’s important to design that service really well, really carefully, with all stakeholders involved. Clinicians must be enabled and given the autonomy to design that service together with the end users – it’s critical that those end users get a say.

We’re working to integrate Hospital at Home with the new EPR system at present. We’re also using a digital solution through which we can monitor vital signs remotely; the patients are receiving care from multi-disciplinary teams in the same way that they would if they were on a physical ward.

Moving forward, we want to really advance this service in terms of the technology that we are using and the infrastructure too. We want to be able to get readings from different devices and to include more patients, focusing on inclusivity. We know that not all patients will have access or skills around digital devices and it’s important that we do not exclude them. Health inequalities assessments when designing new processes are critical.

Essentially, we need to get the data, analyse the data, learn from it, and refine services even further.

Patient flow and virtual care across the Humber

Lee: We’re in a similar place. We’ve got a good penetration of care homes for the Summary Care Record which is a main focus at the moment, particularly looking at who is going to be using those technologies, and there’s work happening around patient-held records too. We’re also doing a lot of work with the Federated Data Platform. We’re using a dashboard to pull the data together from a visual point of view for multi-disciplinary teams, which we wouldn’t necessarily get from the Yorkshire and Humber Shared Care Record as that takes a more individual focus.

Although we’ve got a number of different layers there, with electronic patient records across a number of providers, as Penny said it’s about designing care around the person. You have to decide how to provide care in a virtual sense as suits the population, and it can be slightly different in terms of where you provide step-down care and what that step-down care will be. For example, in some areas, we have community hospitals to provide step-down care as well as virtual wards.

Because you’ve got multiple providers and multiple systems including local authority, pulling information together is a challenge. Not sharing data, not making it available, is the worst thing we can do, for the patient and for professionals. You need it for integrated care planning. In some ways, we’re very joined up – our local authorities have access to our health information, for instance, but we do have some gaps where it’s just a case of getting data of the right kind and quality between different people so that it becomes meaningful.

We’ve got to make sure that this virtual journey is a healthcare and a quality improvement journey, and that we use the right tech at the right time.

Alcidion’s learnings on what matters the most when implementing patient flow and virtual care

Paul: So often, we see a world where people over-focus on technology and functionality and under-focus on people, whether that’s making sure that the technology suits the system users or ensuring that it delivers tangible clinical benefit. When we’re implementing this tech, we have to tether it back to the critical factors – the people and the outcomes.

From our perspective, we need to break down what we mean by patient flow. Are we talking about organisational flow, movement within a structure, or are we talking about flow across a geographical footprint or a pathway where we know that clinical care doesn’t adhere by the boundaries of a setting? It’s vital that we’re clear about what we mean by flow, and also clear about our ambitions. If we want to achieve the ambition of managing flow then we need to start looking at the things that Lee has been leading in his area, with regards to unlocking data and moving it across systems.

Once we’ve defined these things, you need to define the technology that you need to ensure that it’s powered by real-time or near-time data. It’s got to be data standards driven too – if it’s not, you start creating black holes in data and user confidence dips. They start stepping away because the data is incomplete and it doesn’t really help them make a clinical decision.

Alcidion has deployed flow systems with virtual care in Australia, New Zealand and the UK. From implementing these systems, we’ve observed that it really is key to engage users. It has to offer them some sort of gain, a reason to use it. No matter how fantastic a flow system or virtual care system is, if it’s not adopted – if users aren’t engaged and inputting data into it to feed the flow, the automation or algorithms – then it’s not going to unlock benefits for anybody. We’ve seen system adoption work best through really early user engagement, so that user problems are understood and marginal gains are apparent to them. That could mean displaying them in an interactive dashboard, for example, to help clinical teams with buy-in. That, in turn, drives up data quality. When you start delivering the benefits, clinical teams see the difference, and it starts to perpetuate itself.

Penny: It’s worth asking what we mean by engagement, too – it’s not just about asking people what they think about the final product, it’s about walking through their experiences of services. We need to identify the pain points from the perspective of the patients and the clinicians.

Improving patient flow: where to start

Penny: It’s a collaborative, iterative process. It’s about how we improve our processes and services to add value to the people we are serving. Start with that vision. You need to think about your population needs, who you are going to partner with, understand your data. Then you build a workflow or a series of workflows which are integrated across the system. Starting small is key – piloting a service, using well-selected technology and good guidelines and protocols, having good clinical and operational leadership as well as champions to help move it forward. Then you can refine the model that you have created through measuring the outcomes, and scale it to more specialties.

You need to make sure you’ve got good inclusion and exclusion criteria, as it’s not a service that will suit everybody. It’s not just a case of somebody who doesn’t want a digital solution – they might be happy with a digital solution but lack the skills or confidence, so we need to be able to help them with that. It’s a similar situation with our workforce; upskilling is vital.

Lee: I think it comes back to identifying what patient flow actually means to you and the problems you are trying to achieve. You need to look at things like the number of incidents that have occurred during provision of care – they are the areas that you should probably focus on. You need evidenced problems, not just a gut feel.

Then start to identify gaps or hand-offs in the flow of patient care, and information that supports it. That could be a case of something not linking into the system, or a failing in the capability to staff, for example. That’s hard information we can go back to.

You need to know your processes, even though they might not be great. There might be all sorts of workarounds, formal and informal. Those ways of working will beat the technology every time, because culture always beats strategy and delivery. So the key bit is to influence the culture change.

Accept failure. But learn from it, and move fast. Don’t spend two years coming up with a perfect solution that very swiftly becomes out-of-date. Take a lean approach; do rapid prototyping in whatever way works for you to figure out if that solution or process will work before you start putting code down, or set policies, or put workflows in place.

It comes back to the issue around accessibility, too – when we realised that some of our patients were going to struggle with smartphones or tablets, we decided to roll out Alexa with some simple development around the back-end for reminders and alerts. We were able to do that because we walked through the process of how the pathway would work in an environment where someone struggled with digital skills.

Finally, it must be operationally and clinically led. It cannot be led by digital people – they can give really good advice and examples and really help you, but it’s got to be led by the people that will be using it. They don’t need to be massively senior; sometimes a really good administrator in a team can be as powerful as someone in a more senior role. They just need to have respect, to be seen as someone who can set the scene. A medical director, whilst more senior, might never actually be involved in this pathway – although it’s worth saying that they would make a good sponsor for the work.

Paul: Supported by the right core technology and strong governance, virtually delivered healthcare are going to be integral and permanent elements of the broader healthcare system. Critically, it’s going to be about how the data can bridge between the fragmented settings we have at present, break down those service siloes and really start to support how we refine healthcare pathways across organisations.

When you have this core capability, you can start looking at how to use intelligent health services, how make it smart. How can we use artificial intelligence and automation and modelling to support our frontline clinical teams?

I think the digital push for the future is really going be about the continuum from emergency department systems all the way out to self-care, and the opportunity for healthcare systems is much more than virtual flow systems – it’s about digitising new pathways and reducing the friction that care workers experience.

Lee: Having a good foundation is absolutely key. Focus on the ‘fancy’ technology like AI should never come first – you need a decent, fundamental set of data in place before anything else, as well as the data to create the learning environment around tech like AI. The data’s got to be good quality too. Otherwise you have all those data biases built into digital twinning, your flows, your decision support. Getting the basics right on your information creates so much potential; otherwise you’re going to be spending a lot of money to make some poor decisions.

What does good look like for patient flow?

Penny: It’s important to look at it from a patient perspective. You want a safe service that offers an excellent experience. You want to be sure that you can contact your clinical time at any time whilst being cared for at home. If you deteriorate, you want to know how you are going to communicate with your clinical team remotely. You want a reliable digital system where you know how it works, you trust it to work, and you trust that there are people there to look after you if you need them. Those are the basics of good care.

Of course, different patients have different needs, and a good service has to tailor for those needs. As a patient, I would like to be involved in the design of that service and have a say in how the service can improve. I also want to see the benefits of the service – what are the outcomes, how has it benefitted me, my community, how is public investment being used?

Lee: I’d agree – it’s about having a good quality experience. You hear so much feedback from patients saying that the nurse did a good job, for example, but under constraints. We know, within the system, that we cause a lot of those constraints. Our existing workflows are the reason that a patient might have to drive to an outpatient clinic for a five-minute appointment – that journey could take two hours out of their day, it could have given them childcare problems, transport issues, takes money, and potentially much more.

I’d say good is when it is clear who is providing your care, who they are, what they are doing, and you have an input into provision of care. The patient is the most knowledgeable person in terms of what’s going on with them – their carer too, potentially. In some ways, this doesn’t happen enough at present. Patients aren’t sure if they’re being seen by a nurse, a healthcare assistant or a social worker, and a lot of the time, they’re just going with the flow. That’s certainly true within the deprived areas that I know within Leeds, Hull and Bradford for example.

The other side is being able to provide that care and support in the right place. It could be with the family in the home, or it could be in the hospital, but it needs to be considered. The location in which care is provided needs to be made as functional as possible.

We need to acknowledge that it’s really difficult for patients at the moment, with how disjointed services are. It should be as slick and easy as booking a holiday or ordering online shopping, ideally.

Paul: Clinical acuity needs to be understood and monitored, there needs to be visibility. Irrespective of where a patient presents into a system, we need to be sure that the key information about that patient and their patient journey is understood by the relevant people.

If we get this right and we deliver these services as they are intended, a person receives clinically effective care in a setting that matches their need and preference, with a great experience. I think success in this area will really empower people with regards to how they access care, access information, and seek escalation in their care journey when needed.

Essentially, it’s about focusing our services and shifting the paradigm around technology for organisations to technology for people, and aligning to that vision.

Technical considerations 

Paul: From my perspective, data standards are critical. They’re not quite technical, but they do inform the APIs and the behaviours, the integration points around how we manage the movement of information. I’d recommend maintaining data dictionaries to ensure that we’ve got clarity on the information that is being moved around.

Ultimately, we need to think of it from a solution perspective, because it really does require the contribution of that maze of wider organisations.

Lee: I’d agree, standards are absolutely key. We don’t realise the standards that exist outside of technology, we just assume that they are there. The challenge in healthcare is that we do have a number of standards, but they’re not necessarily applied. It’s simple when you think about it – the reason that we can put petrol into our cars is because it’s built to a standard, as long as you’re putting petrol in a petrol car, it will work. For data sharing to work, we need to focus on those standards.

There are gaps in those standards however – though I would say we shouldn’t let perfection be the enemy of good on some of these issues.

Vendor relationships are also key. We need to develop partnerships with our vendors and work together to get the best outcome. Realistically, everybody is trying to achieve the same thing. We need to understand each other and have clear goals in place with regards to how we work together. In most cases, for information flow, you’re pulling information from a number of different solutions and different vendors. They might have slightly different objectives or reasons but I do believe that our vendors are in this because they want to improve patient care. Otherwise they would be in banking rather than health.

A lot of the challenges around technology and standards come back to legacy in terms of how the NHS is reconfigured. There are frequent changes and the NHS can lack in direction. Realistically, that’s where the opportunity around ICSs comes in. The Hewitt review hones in on this – there’s a strong focus on giving the ICS the responsibility to deliver. When we talk about patient flow, ultimately we’re talking about all the different systems talking to each other and focusing on the individual – professionals having a good experience of the system so that they actually use it, and the system itself adding value. It’s about giving that responsibility to a footprint that is large enough to really have a focus. When we’ve got that, we can focus on the place for delivery.

Paul: Lee has hit on something there that is so true and I just want to emphasise it from our perspective, around organisational behaviour in partnerships. You could say that integration data flows and patient flows are not technical challenges any more – largely speaking, they are behavioural challenges. It’s so important that we work with organisations and partners that have that shared purpose and willingness to have open APIs and engage in this work collectively.

The thing that I’ve seen be most influential on success is behaviour and relationships, as opposed to technical barriers. Technically, there are always solutions you can find and adaptions you can make. But fundamentally, if organisations don’t have that integrated open collaborative approach, it’s very challenging to overcome.

Lee: The reality is, we’ve all got to work together for the same aims; the challenge is making sure that your aims are all aligned in the same way. We have to accept that there will be a bit of variation between different ICSs or organisations, but fundamentally, we are all here for patient care.

Many thanks to Penny, Lee and Paul for joining us and sharing their thoughts.