This article was written by Sarah Ouanhnon, Senior Consultant at GE Healthcare Partners. Previous articles in this series: The next wave of Healthcare Innovation will be driven by Primary Care and Surely it's time for primary care to work differently

anuary 2019 

I have had the privilege of working on various digital solutions for healthcare, from e-monitoring and care coordination websites to decision support tools, workforce scheduling systems, and population health platforms in both the UK and France.

One question that is consistently on my mind is – who do we build these tools for? This is a particularly critical question given the many announcements in the NHS long term plan and the latest Stratégie nationale de santé about digitising the care management process – from appointment booking to GP Consultations. Everywhere you go, everyone is talking digital.

But who should we be building the digital tools for? Do we build them for the patients or for the clinicians, for the regulators or for operational managers?

In my view, the answer should always be the same: build it for the user. The benefits for the other stakeholders will come naturally, as long as the specifications are comprehensive enough from the beginning.

This is a general observation in healthcare, but it is particularly true in Primary Care where users were often more isolated, making them more difficult to reach. However, with care moving closer to home, the need for more coordination, and the drive for prevention, it is urgent to design digital solutions adapted to Primary Care.

So, what have I learned along the way?

  1. Figure out who the user is

    Everyone agrees that we should always focus on the user, but the question is – who is the user? In a field as complicated as healthcare with so many participants in the care continuum, the first and most critical challenge is figuring out who the actual user is in a particular scenario. Some approaches take a default position that the user is always the patient – but that is not true. Take the case of workforce scheduling system. Whilst the overall objective is to maximise the effectiveness of the service in meeting the needs of patients, the real user here is an operations manager sat in a GP practice trying to figure out how they can best deploy the team to provide the capacity they need – including balancing so many conflicting demands. If we design a solution for them, then we make a huge difference. Notice also that I did not say, the Ops Manager – as finding the user is not about finding the person with the ‘job title’, it is about understanding the specific task that needs doing and where in the service that task currently sits. That way, we create more effective tools – generating lasting value for both the service provider and the patient.  Sounds obvious? The reality suggests otherwise. Many systems continue to emerge that have no focus on the ‘user’.

  2. Designing the right features: ask the user what they need

    Most major changes in the field of digital solutions for Primary Care are driven by regulatory change. For example, in France, the impact of the certification of prescription tools for Primary Care or more recently the national “label” for Multidisplinary Primary Care practice softwares1. These changes are valuable drivers both for the software providers and for the Primary Care users, but they cannot be enough in defining what the users need to support and enhance their role. Another example is the change driven by new payment mechanisms: GPs in France must report several indicators on their practice to get additional yearly payment (“ROSP”). Most reporting modules have therefore been designed with the sole objective of creating quick and easy reports to be transmitted to the National Insurance Services, without takinlab2690g the time to think about how these functionalities could actually be used for improving practices, or even how to make sure that the data is accurate. 

    Another common pitfall is to build solutions with the specialists only. To develop a Primary Care functionality for Asthma, we ask the pulmonologists, to develop a new formulary on Women’s Health, we ask the gynaecologists. Their input is obviously important and should be considered but the final design can only be done with Primary Care practitioners. An interesting illustration that GPs have shared with me is the temptation of “alerts” and “reminders”. For a specialist focusing on one pathology or one organ, it could seem useful to set up automatic reminders in the electronic record – for history checks, vaccines, or screening tests for example. But the major risk is what is called “alert fatigue” which could result in the user not paying any attention or, worse, simply disabling this nonetheless useful information! How can we make this type of automation relevant to a Primary Care context addressing multiple pathologies? Only Primary Care experts could help us answer that.

  3. Designing the right interface: ask the user how they work

    A simple but very important point is that interfaces should be appealing and intuitive. What is the reaction of anyone today suddenly browsing on an “old-school” looking website? Most of the time, we just close it. Why does it have to be different in Healthcare? I have heard many times software providers say that Healthcare staff require training in how to use the solution. But who needs training to use Gmail, Facebook or an iPhone? Why should Healthcare specialists need to receive training before using any solution? Yes, Healthcare is complex and there will always be some functionality that needs more advanced training. But intuitiveness of a solution also comes from the way it has been developed (working closely with the end user) and basic features should be easy to access and understand with no more than basic peer training.

    Historically, IT systems were designed with the back-end and database first and the user interface only at the latest stages. Today’s world is different. The main differentiator of a Healthcare IT solution is in the ease of use of the user interface. Talk to clinicians about their everyday software and they can tell you how many “clicks” they need to get to such and such functionality. Don’t get me wrong, the back-end is still crucial and ensuring it is architectured well is critical for the longevity and perAppformance of the system. But it shouldn’t be the starting point. Starting with the interface ensures not only that the solution is focused on its user but also that your specifications are comprehensive.

    A good example of this is the implementation of “teamwork” functionalities. In the context of Primary Care, teamwork need goes beyond being able to see the same screens. It covers information sharing (relevant information only) but also coordination tools. No one can define the requirements of these functionalities and write specifications at their desk. It is necessary to observe, spend time with users, understand what they currently use papers or ad-hoc solutions for, what the ‘pain points’ are and what the opportunities for improvement could be. And test and re-test the user interface with mock data even before starting to develop the back-end.

    We need to accept failure and learn from it, in order to get a really adapted solution as quickly as possible. The idea of ‘rapid prototyping’ (that we find in the Design Thinking approach for example2) is helpful here. There is no need to have a fully functional solution to start usability testing. This should also be the approach with digital solutions for Primary Care.

  4. The right innovation: ask the user what they will do tomorrow

    The near future is likely to see major developments in Precision Health, Robotics, and Artificial Intelligence, all of which are really promising for Healthcare. But technology should not be the only driver for innovation. While researchers focus on these advanced techniques, the most urgent issues that our Healthcare Systems struggle with today are organisation, coordination and efficiency. Clinicians lack time and resource because of this – repetitive admin tasks waste significant time. How can software solutions address this? By helping predict a patient’s diagnostic? Probably, in the future. But there are easier and more immediate answers. Automation of treatment protocols, prescriptions, embedding of guidelines and best practices: all these features are far from fully developed in the existing solutions whereas they could be entirely built using only current technology.

    3011715 CTEM0R2Spending time in Primary Care practices showed me how difficult it is to have continuity of information on a patient. I already mentioned the difficulty of coordination within a practice, but it is nothing compared to the difficulty of coordination outside the practice. Digital technology could be a great asset in improving information sharing and continuity. But the reality today is that as soon as patients are referred, we lose track, until one day they come out of the hospital, knocking on the GP’s door for advice following a hospital stay, a hospital stay that the GP was never aware of. Far from the latest e-health tools and the wonders of telemedicine, digital solutions still do not support basic coordination between care providers. This is the type of innovation that our systems really need, not a technological innovation but a design innovation to make the best use of the technology we already have. And this does not diminish the challenge! How can we capture and understand how professionals will collaborate tomorrow across care settings? How can we design for so many different users at the same time? And how can we maybe involve and empower the patients in some of these coordination tools? Innovative ideas and approaches are clearly needed.

    A simple way to prepare for tomorrow’s predictive algorithms, while bringing value today, would be to give a new focus to data. And I’m not talking about trendy “big data”. Working with Primary Care specialists made me realise how little visibility they have on their own work. Apart from payment-related dashboards, it is very difficult for Primary Care teams to know the impact of their interventions on patients, let alone to have relevant information to implement preventive actions and a Population Health approach. However, this is exactly where the future lies – in real-world data used to support a proactive rather than reactive system and continuous improvement. How do we make data collection and structuration less of a burden? How do we make analysis and reporting easier and more insightful? By giving Primary Care teams simple and visually-appealing reporting tools, we empower them in their practice today while improving the quality of the data that can be used for Machine Learning tomorrow!

All these hours spent discussing, observing, testing taught me one simple concept: ask the user! And to go further and apply some more Design Thinking methodologies, the best way forward should be co-designing digital solutions. There will always be technical constraints and therefore a critical role for engineers but this should always be done in co-creation with end-users, with professionals using the solution every day. The next question is then: who should I ask? Once I have identified the user, how can I make sure that I have a representative user or group of users? This is not an easy one. Maybe the topic for a future article?

Primary Care is the backbone of our health & care system. By some estimates approximately 85% of all patient contact across the NHS in the UK, occurs in Primary Care. Similar statistics abound across the globe. At GE Healthcare Partners, we believe that genuinely transforming health outcomes for a population requires a robust and resilient Primary Care system that is aligned to the needs of the population; integrated fully into the care pathway and empowered with the support, tools & resources it needs to improve people’s lives and help the system save time & money.

We also know that Primary Care is full of clinicians who are innovative, entrepreneurial and/or have insights to share. In this forthcoming series of articles we want to provide a platform for those who have visions for what could be different in healthcare leveraging technology and what can be done to accelerate the diffusion into mainstream services.


[2] For more on this, see for example Design Thinking ( and Why You Should Talk Less and Do More (

> <

Share This: