You are here

Data Science: Q&A with IFoA Member Lisa Balboa

Lisa Balboa chairs the Regulation & Ethics work stream of the IFoA’s Data Science MIG, and is an active contributor to the IFoA’s Health and Care Practice area. Opinions expressed here are her own.

 

Lisa BalboaTo what extent do actuaries already use data science tools in the healthcare sector? And from your perspective, what are the benefits of establishing transferable skills between data science and actuarial science?

In healthcare, one of the most impressive uses of data science I’ve come across is hospitals using analytics to forecast the number of patients likely to arrive at their doors in the coming week. Hospital managers use these predictions to adjust staffing and other resources to ensure anticipated patient demand is met. This forecasting is particularly useful in times of peak demand, such as ensuring hospitals have the capacity to cope in a severe winter flu outbreak.

But these models are only being used in a few isolated hospitals. As actuaries, we’re in a good position to draw on our understanding of these data science techniques to help scale-up and encourage more widespread adoption of predictive healthcare tools. We can also use our strong skills of communicating technical information to non-technical stakeholders. This supports healthcare decision-makers to understand, and react appropriately, to these dynamic healthcare forecasts.

What kind of framework must be in place for data to be readily exchangeable across health system stakeholders? And what potential exists to scale-up opportunities in using this data for predictive healthcare demand modelling?

Having common data storage and programming languages is vital if we’re to succeed in embedding predictive healthcare demand modelling right across the health system. Initiatives, such as the FHIR (Fast Healthcare Interoperability Resources) standard, are a great first step in moving towards a system where data can be exchanged across the health service. It’s important to have robust data privacy and data governance frameworks that empower individuals to engage with their health and lifestyle data.

I envisage a personal health record system that places individuals at the centre of their data. The personal health record would allow each person to take ownership of their data and decide whether and when to share this data with their medical team. Individuals could also give consent for the de-identification of their data to be fed into medical research and healthcare resource management activities.

As adoption of AI-based data science tools in healthcare increases, how important is it to ensure that algorithms can be embedded into society in a fair, ethical way, to identify and mitigate possible algorithmic biases? And what can actuaries contribute to public debate around this topic?

The adoption of AI-based data science tools is on the rise throughout pretty much all aspects of our daily lives. With algorithmic decision-making becoming so pervasive, it’s important to ensure we embed this technology into society in a fair and ethical way. Actuaries have much to offer in this area. From a technical standpoint, we can help to identify and mitigate biases that are present in training datasets being used by the models. We also have a lot to contribute when it comes to developing these algorithms. Many such algorithms are designed to maximise resource efficiency. But we should be going further to consider fairness as part of the resource allocation process.

In a field with such a rapid pace of development, ethical adoption of new data science techniques is critical. Particularly as regulation can often take time to catch-up to technological change. As actuaries, we’re well-situated to engage directly with the public and other stakeholders, to help them to understand how their data is being used by these algorithms. We can help to ensure that algorithmic decision-making is transparent and explainable. This will give the public greater opportunity to voice their opinions on what they believe is a fair and ethical way for this pervasive technology to be embedded into our lives. 

What career opportunities for actuaries to work alongside data science specialists in the healthcare sector do you see – now and going forward?

The knowledge and expertise that data science experts bring when analysing large volumes of specialised medical data helps to drive developments in healthcare analytics. As actuaries, we are likely to remain highly reliant on data science experts for these deep insights into specialised datasets. Also as actuaries, we are well-placed to work closely with data science specialists by combining these healthcare analytics insights with our strong understanding of business, finance and risk management.

This could lead to growing career opportunities for actuaries to work alongside data science experts to build models that optimise financial and medical resources. I’d like to see actuaries and data science specialists collaborating with a wide range of other stakeholders such as healthcare professionals and medical researchers to help develop more effective, personalised treatment pathways and thereby improve health outcomes for society.

With specific reference to insurance, how do you believe data science techniques will prove of the greatest value to actuaries working in this field of expertise?

We’ve only just begun to scratch the surface in terms of thinking about how to embed the wide array of emerging data science techniques into our actuarial work. There’s tremendous potential to drive value creation by applying these tools to improve our understanding of how risks interact and to optimise insurance portfolios. It’s interesting to consider this question in the context of broader technological change. We’re increasingly seeing consumers adopting wearable tech, mobile apps and other connected devices into their daily lives. This technology is helping people to engage with their health more often and in new ways. With strong data privacy and data governance frameworks in place, there may be the opportunity for actuaries to use data science techniques to gain a better understanding of customers’ key insurance needs.

Using this deeper understanding of the customer, actuaries could champion the development of new insurance products that are more dynamic in their response to individuals’ health and lifestyle needs. With a growing availability of data and with technological developments continuing to raise customer expectations, actuaries will increasingly adopt more advanced data science techniques. I’d like to see us collaborating with a wide range of experts to embed data science into the insurance industry in a way that enhances business value and drives improvements for customers and society.

Filter or search events

Start date
E.g., 11/07/2020
End date
E.g., 11/07/2020

Events calendar

  • Spaces available

    This webinar is intended to raise awareness of the shifting landscape of climate liability risk and what it means for actuaries, including how it impacts on their professional and legal duties. Presentations will cover the legal risks around climate change for investment consultants and actuaries advising DB pension schemes as well as consideration of climate risk for insurers.

  • Spaces available

    Climate change risks are likely to become material for many risk management and investment decisions. This will require to incorporate explicitly climate change in the tools used for risk management and investment decisions. At present existing climate change tools are often too crude for decision making.

  • Current Issues in Life Assurance (CILA) Webinar series

    Webinar Series
    15 July 2020 - 3 August 2020

    Spaces available

    CILA is one of the pre-eminent events in the annual 'Life' calendar. Due to COVID-19 we are running the programme as a series of webinars covering topics aimed at practicing life actuaries from life offices, consulting firms and other employers of actuaries and those who work in or advise on, the life assurance market in the UK and Europe.

  • Spaces available

    Current Issues in Life Assurance

    For annuity writers, a key challenge is the need to fund capital-consumptive new business strain (NBS) as a consequence of writing the business intended to fund future distributions.

    Reinsurance, investment strategy and capital provision all have roles to play. Here, we:

  • Spaces available

    Current Issues in Life Assurance

    Mortality in 2020 is now dominated by one thing, although – in our future-focused world – the pandemic is just one of many mortality considerations.  In this session, three well-regarded mortality/longevity specialists provide an overview of:

  • Spaces available

    Because of Covid-19, forecasters predict a severe recession in 2020, followed by a V or U-shaped recovery. This impacts both individuals and companies. However, compared to previous recessions, the impact on banks of higher credit losses should be mitigated to some extent by government actions. 

  • Spaces available

    Part of the Protection, Health and Care Conference 2020 webinar series

    This session will provide an overview of the Population Health Management Working Party's research including defining impactability and impactability modelling, discussing some examples of specific modelling approaches, considering the practical challenges across the NHS as well as wider public perception and ethical issues.

  • Spaces available

    Many actuaries consider career opportunities in the Finance and Investment practice area after having started off in more traditional actuarial roles such as valuations, capital management or pricing. This session is aimed at helping actuaries to better understand roles in Finance and Investment and how they can fine tune their skills to pursue such careers.

  • Mortality and Longevity Webinar Series 2020

    Webinar Series
    22 July 2020 - 10 August 2020

    Spaces available

    Due to COVID-19, we are running this programme via a series of webinars commencing 22nd July.

    This webinar series will provide topical and practical updates and discussion on the latest thinking and innovations in mortality and longevity, and is designed to be very accessible to a broad range of experience.

     

  • Spaces available

    Predictive risk assessment and risk stratification models based on postcode-level consumer classification are widely used for life insurance underwriting. However, these are socio-economic models not directly related to health information. Similar to precision medicine, precision life insurance should aim to tailor policy pricing/reserving to the individual health characteristics of each client.

  • Spaces available

    As insurers look towards their internal model calibration process for 2020 final year financials and statutory returns, actuaries need to deal with the complexity of adequately modelling their business in 2020 and beyond. This discussion will look at what poor model selection and calibration could look like – using inappropriate historical data; using incorrect 2020 mortality data; inappropriate stochastic model recalibration (or lack thereof). What about being prudent vs setting a best estimate? How do you allow for tail risks during a tail risk event?  This is the fourth webinar in the Extreme Mortality Events series presented by Chair of the Life Board of the Institute and Faculty of Actuaries, Colin Dutkiewicz. 

  • Spaces available

    This webinar has been re-scheduled from its original date of the 1st July. Although ESG has many buyers across the asset allocation community, from pension funds to sovereign wealth funds, it still hasn’t found its place within the core asset management strategy desks where the money is actually invested. The problem as well as the opportunity is Fixed Income. Plenty of strategies exist for incorporating ESG within Equities, from screening, integration to a combination. ESG has picked up relatively quickly within Equities with rating,indices created using ESG factors. This talk will discuss how we price a quantifiable ESG credit risk premium and make it alpha worthy in a strategy. 

  • Spaces available

    Part of the Protection, Health and Care Conference 2020 webinar series

    With the rising prevalence of dementia, how can we manage this risk effectively and can insurance do more? Matt Singleton, Ageing Lead at Swiss Re, will cover these topics and demonstrate how insurance could help people address their concerns.

  • Spaces available

    Current Issues in Life Assurance

    This talk will look at a range of such techniques (e.g. mass lapse risk transfer, contract boundaries, risk margin relief, non-standard longevity risk transfer) that have been applied or considered by UK and EU insurers, and the pros and cons of each.

     

  • Spaces available

    Current Issues in Life Assurance.

    The International Association of Insurance Supervisors announced on 14 November 2019 the adoption of v2.0 of the global Insurance Capital Standard (ICS) which will undergo confidential reporting for 5 years starting from 2020. This session will include specific experiences from Legal and General (L&G) as well as global industry perspectives from EY.

  • Spaces available

    Current Issues in Life Assurance

    This session will cover the PRA supervisory statement on financial impacts related to climate change, industry insights into PRA climate risk business plans, examples climate risk strategy setting out key workstreams and activity steps for successful execution, an overview of a climate risk strategy execution timeline and the future.     

  • Spaces available

    Part of the Protection, Health and Care Conference 2020 webinar series. Using new and unique research and data from the UK, US, Sweden and China, this presentation investigates how consumers use the internet through their insurance journey and analyzes the role culture and generation plays in their online behaviour. We use this research to show the online landscape for insurance sales in the UK and suggest ways to shape new products and effectively engage with the consumer who is buying them.

  • Spaces available

    Chief Medical Officer (CMO) for Gen Re Life/Health Research and Development, Dr John O'Brien, will discuss the impacts of Gene Modification for life/health insurance. 

  • Spaces available

    As an industry, it has been important to be able to look to the future to identify the next quantifiable risk. In this session, I will explore some of the less tangible, but none-the-less concerning risks to future health, such as the health risks associated with exposure to pesticides, ingestion of plastic in the food chain, and the hazards of indoor air pollution through exposure to volatile organic compounds.

  • Spaces available

    The working party will help the industry to update and enhance how potential risk from diabetes and excess mortality is considered, including the need to understand the underwriting implications as treatments improve, and potentially to develop new products that are tailored to those with diabetes.

  • Spaces available

    Part of the Protection, Health and Care Conference 2020 webinar series. Modelling the structure and trends of cancer morbidity risk is important for pricing and reserving in related health insurance fields such as critical illness insurance and care provision. We model the dynamics of cancer incidence over time in different regions in England, using 1981-2016 ONS data. The modelling allows estimation of cancer rates at various age, year, gender and region levels, following a Bayesian setting to account for statistical uncertainty. Our analysis indicates significant regional variation in cancer incidence rates. 

  • Spaces available

    Part of the Protection, Health and Care Conference 2020 webinar series. In this talk we will outline the steps Aviva took in pulling together our first large-scale disclosures on the exposure of our business to climate change published in March 2019; in line with the recommendations of the Taskforce on Climate-related Financial Disclosures. After touching on why insurers have such an important role in climate change, we'll cover a brief “how-to” guide for those who have not yet embarked on thinking about these topics before giving a case study of how the learnings from a TCFD disclosure exercise can be applied to investment portfolios.

  • Spaces available

    Part of the Protection, Health and Care Conference 2020 webinar series. 

    The insurance industry currently underwrites customers with diabetes based on a range of factors, medical expertise and various medical studies. The work undertaken by the Diabetes Working Party would help the industry to approach this using current research findings to update and enhance how potential risk from diabetes is considered. This includes the need to understand the underwriting implications as treatments improve, and potentially to develop new products that are tailored to those with diabetes. This webinar will present our latest findings in the management of this important chronic condition which will include research in collaboration with the ARC.