A Conversation with Pascal Mwele
“The connection between innovation and health outcomes is very clear to us.”
Pascal Mwele joined Palladium in January 2019 to lead the Centers for Disease Control (CDC)-funded Kenya Health Management Information System (KeHMIS) II project as Chief of Party, and transitioned into a global role in 2023 as the Director of Innovation and Change Management, based in Palladium’s Nairobi office.
Pascal was appointed Technical Director for the Data for Implementation (Data.FI) project in May 2023, a role that entails leading technical implementation across the project’s national and regional programs, as well as serving as a vital thought leader for the project’s digital solutions. He has more than 15 years of experience leading and supporting national and international technology projects in sub-Saharan Africa. After two months in his new position, he agreed to share his professional story and perspective on the wider digital health sector.
Good morning, Pascal. What drew you to the digital health sector?
I have a background in information system management and software development; however, my undergraduate degree is in computer science. In my final year of undergrad, I developed an interoperability messaging service to trace the movement of antiretroviral drugs. My choice was based upon living with a relative who was living with HIV. Later, I learned there were lifesaving drugs that would often run out of stock and lead to drug resistance. I became convinced that technology could play a role in ensuring there were no stockouts and could solve other intractable challenges in low- and middle-income countries like Kenya.
What did you do before starting your career in digital health?
My experience spans both the private and public sectors. For example, my first big project was in utilities, and before that I developed microfinance software. I was exposed to technology in different fields, and in organizations large and small, which allowed me to see technology's potential in different contexts. Consequently, I was passionate about understanding how digital technology could be applied to health. Those experiences, along with my earlier personal experiences, got me interested in digital health.
Could you speak about one moment from these earlier positions that informs the work you do at Data.FI and Palladium?
If I had to point to one specific experience, it would involve developing a solution for a Nairobi-based utility company. I was hired as a software developer and quickly realized that I could not apply technology to the front lines without interacting with front-line users, experiencing the use of technology (or lack thereof), and understanding what the biggest problems were on the front line. That experience was a light bulb moment for me; it opened my eyes to the reality that being a 'techie' is not synonymous with locking yourself up in a room and churning out code.
“Any project or initiative related to technological transformation must always put people at the center.”
Instead, I learned that I should always put people first, that I should always strive to experience the challenges that your users face. Any project or initiative related to technological transformation must always put people at the center. As in the private sector, so in the public sector.
And how do you define your role now? Is it more about maintaining continuity or identifying new opportunities?
My responsibility as a leader is to do both in a responsible manner.
First and foremost, I want to make sure that what works well continues to work well. As one of my previous project teams put it: 'Continuous improvement is better than delayed perfection.' It's the same mindset I'm bringing to the team – improving on what's working while also seeking out new opportunities. A 'Big Bang' change is often difficult to pull off because transformation is often disruptive, so facilitative leadership can never be overemphasized in leading a team to their goal. The importance of contextual awareness in change management cannot be overstated; the ability to work within a culture, especially when it has delivered good results before, is extremely important. Culture, as they say, eats strategy for breakfast.
Speaking of contextual awareness, how has the sector changed since you first got involved? What are the new challenges?
Initially, funders, investors, stakeholders, and user communities considered technological solutions as a 'nice to have,' not a 'must have.' As a result, technology was an administrative tool, not necessarily a strategic resource.
Today, most people working in digital health deal with 'digital natives,' people who understand the value of technology and how to use it effectively. Therefore, stakeholders now expect a lot more from technology. This often leads to a disconnect between what is expected of digital health professionals and what is actually practicable or useful.
It is not a new challenge, but it requires better people management skills, continuous knowledge acquisition, and upskilling of both digital health professionals and stakeholders.
Can you describe an example from a country you've worked in where that disconnect has been overcome?
I have been privileged to have been part of quite successful projects. That said, it is impossible to always get it right all the time. Often, teams develop a very good solution only to go out in the field and discover that the challenges that people are facing are sometimes very, very different from the solution that you are looking to provide.
For example, in a past engagement, we were trying to develop a cross-border solution that allowed mothers to carry their children’s immunization records electronically when traveling across international borders. Initially, we thought about a smart card solution, so we engaged potential end users together with our partner to look at the feasibility of such a solution, and found that even power access was a challenge, let alone supporting smart card infrastructure. Also, the number of mothers coming to clinics was higher than we expected, and this put immense pressure on healthcare workers. There was an opportunity for automation, but the approach had to be carefully thought through.
It was only after engaging with the people in the field who were going to interact with this technology, that we realized we needed a solution that is very low tech but also has a high level of utility. This confirmed the importance of reaching out to the user communities very early, testing your thesis statement early, and continuing to refine the solution with the users as you evolve and develop technologies.
In this case, we ended up with a solution that was based on QR codes, allowing records to be stored and retrieved safely in a way that did not compromise data safety requirements.
What innovation(s) in our field are you most optimistic about?
Machine learning and AI, no doubt. Machine learning to power artificial intelligence, or augmented intelligence, are here to stay, and I'm very excited about what Data.FI is doing to embed machine learning in natural and intuitive ways to support augmented intelligence. In the same breath, I am very proud of what my KeHMIS team did to develop and deploy machine learning models at scale and in very low-resource settings, in terms of challenging access power, internet connectivity, and computing resources. I am eager to transfer lessons to what Data.FI is already making progress in.
For example, most countries facing HIV epidemic control struggle to find the next positive person and prevent the subsequent positive case from occurring. Thus, we adapt machine learning models in a way that allows healthcare practitioners to identify people who are likely to become HIV-positive if no intervention is applied immediately. Of course, this must be done ethically, while respecting the clients’ privacy and security rights, explaining the drivers of model outputs transparently, and always deferring to the healthcare worker’s judgment.
The acceptance of these kinds of techniques and approaches has increased, particularly when machine learning is ethically safe, transparent, nondisruptive, and can be implemented in very low-resource settings.
In digital health in particular, machine learning can be deployed in a very sustainable and intuitive manner and can be a powerful tool to power person-centered self-care for promotive and preventive primary health care programs.
I always think of the quote that ‘AI will not replace an individual or an organization, but an individual or organization that uses AI will definitely replace one that doesn’t.’ The organizations that can quickly pivot into applying technology while remaining people-centered and contextually aware will be the breakthrough organizations.
“The organizations that can quickly pivot into applying technology while remaining people-centered and contextually aware will be the breakthrough organizations.”
In contrast, what global health challenge do you think people should pay more attention to?
Pandemic preparedness has become the focal global health challenge due to the unprecedented global COVID-19 pandemic. Awareness and investment have ensured that robust systems are in place to predict, prevent, detect, and respond to public health crises. However, not as much attention is given to strengthening routine local, national, regional, and global data management systems that would best prepare us for the next wave or outbreak.
This challenge requires all stakeholders to play their part. At Palladium we are keen to not only support digital health transformation, but also strengthen initiatives that enhance health data governance. For instance, in Kenya, we joined other experts and organizations in working closely with the Ministry of Health to adopt the Global Health Data Governance Principles and contextualize them to the Kenyan setting.
Why is this important? Because as we get to technology maturity and deal with the general increase in digital literacy, we have to put the right guardrails around the implementation and design of technology and the data the systems generate. The data subjects, individual users, must be aware of what the data that is collected from them is used for, and must always have the opportunity to opt out. Digital health is now transitioning away from facilitating extractive data collection tools to being enablers of personalized selfcare, as much as they support the healthcare workers to deliver services. With this, there is real risk around the exploitation of health data in unethical ways.
On the other hand, it is important to promote and expand the responsible use of data, because this allows person-centered care and easy identification of potential outbreaks. It also promotes health-seeking behavior from an individual-level perspective and empowers clinicians and healthcare providers to be proactive in providing care to their communities.
How to ensure an equitable and effective balance of these considerations is very difficult. I think at least part of the challenge is in how fast technology development and adoption is taking root. In the health sector, there is an opportunity to learn from other sectors and adopt the frameworks used for digital transformation.
When you look at banking, for instance, the banking sector adopted technology as a strategic business driver quite a while back. So, while you share a lot of personal information with your bank, you most often are not afraid of sharing that information, because you can be reasonably sure that your money, privacy and identity are protected. But that is not as much the case in the health space. The governance and legal environment around health data is not nearly as mature as with the financial services sector.
That’s why we need to support countries to leapfrog, and learn from sectors such as banking, so we don’t have to go through the long learning curve that the banking sector had to go through. The countries that navigate that process responsibly are going to have an edge in digital health innovation.
Why is Data.FI well-positioned to help tackle this challenge?
I think we take a very pragmatic approach in implementation, one that looks at designing technology with the end goal of ensuring that it can scale, and that innovations are designed with the users, for the users. The connection between innovation and health outcomes is very clear to us.
Data.FI is also unique in the sense that we are dedicated to a continuous improvement process. We partner with government agencies to innovate at scale, but we don’t introduce solutions and then walk away. We continue to be dedicated to working with ministries of health, in-country partners, and with funders to hone solutions and governance structures that can withstand evolving contextual priorities and conditions.
Lastly, we are always looking at the long-term utility of the technologies that we develop, support, and deploy. Whether that means that we develop accessible technologies that can be built upon by others, build upon other organization’s work, or when we're developing the technologies together with other organizations, including indigenous organizations, we are always transferring skills to in-country professionals and organizations.
Collectively, I think all this lends itself to positioning us as a differentiator in how we develop and deploy technology in the countries that we're working with.