Deep expertise in modelling non-communicable diseases, combined with entrepreneurial vision

Why us?

HealthLumen is committed to playing our part in helping address the epidemic of non-communicable diseases (NCDs) in close collaboration with our clients.

We bring to the table a highly qualified team and over a decade of experience, all underpinned by our core values:

We are rigorous
We dive into challenges and think them through properly. We’re thoughtful and thorough. Which is why our customers trust us to always do what’s right.

We are inclusive
We bring together epidemiologists, mathematicians, product designers, engineers, data scientists and more. We’re a diverse community with diverse perspectives that assist our customers in helping make the right decisions to improve health outcomes.

We are adaptable
We listen, think and do. We’re open to feedback and adapt to change. We’re receptive, responsive and constructive – with each other and our customers.

And our entire focus is dedicated to addressing the NCD crisis.

The epidemic of NCDs

NCDs including heart disease, obesity, diabetes, respiratory disease and dementia are the leading cause of death worldwide (World Health Organization, 2020 ¹). They also represent a huge economic burden on health systems, governments, employers and health insurers.

The financial burden of 5 major NCDs (cardiovascular disease, chronic respiratory disease, cancer, diabetes, and mental illness) could contribute a cumulative output loss of US$ 47 trillion in the two decades from 2011, representing a loss of 75% of global GDP in 2010 (US$ 63 trillion).¹

To respond effectively, it is crucial to understand these epidemics better:

• How will the burden of disease change over time based on current trends such as demographic change, changes in risk factor prevalence, or changes in disease incidence?
• What effect will different interventions have?
  • Interventions include new drugs aimed at slowing the progress of a disease, digital therapies to help manage a chronic condition, and public health interventions that aim to reduce the key causes of NCDs which include risk factors such as alcohol consumption, smoking, poor diet, physical inactivity and air pollution.

Tackling the NCD epidemic: the role of modelling

Computer-based modelling can help quantify the impact that different inputs, such as changes in behaviours, or new interventions, have on disease morbidity, mortality and costs. NCD modelling is a specific method for estimating the extent to which changes in one or more risk factors affects disease and health.

This then provides evidence to inform decisions, for example, made by payers about which technologies and/or services to fund, or by pharmaceutical companies developing new therapies, or public health bodies considering policy interventions.

To make these decisions effectively, the clinical and economic value of each option must be estimated, taking into account costs and outcomes into the future, for a complete range of population subgroups.

Modelling has two main applications to assist decision-making:

• Estimating future disease trends based on demographic change or predicted changes in risk factors.
• Quantifying the impact of interventions to help understand the effect of prevention, screening or treatment interventions on health and economic outcomes by comparing a population treated with a certain drug or policy intervention compared to how it would develop in the absence of treatment, or if treated by a less effective treatment option.

Modelling and Real World Studies

Modelling studies are cheaper and quicker than real world studies; furthermore, in many cases true randomised experiments are often not practical, cost-effective or ethical.

However, modelling should be seen as a complement to real world studies because they help get the best value out of these studies, often being used to integrate evidence from different studies and different domains.

Modelling can also inform empirical studies, for example to identify assumptions or key parts of evidence that may have a significant impact on model outcome (direct empirical study of these assumptions may give greater confidence in the model and its predictions). Modelling may also be used to estimate likely effect sizes to inform the size of evaluative studies.

What is microsimulation modelling?

Microsimulation models are based on computer-generated “virtual populations” that reproduce the characteristics and behaviour of a large sample of individuals representing the whole population of interest.

Microsimulation represents a heterogeneous population by reproducing the characteristics and behaviour of a large sample of individuals, using the best data available of population demographics, risk factor status, disease incidence, mortality etc and, can overcome gaps in existing knowledge or data.

Major life-course events can be represented in the lives of the simulated individuals and, in the case of dynamic models (such as HealthLumen’s), characteristics and behaviours can evolve over the life course.

• Events like pregnancy and birth, exposure to risk factors like hypertension, cholesterol, smoking, air pollution and changes in weight, can be attached to each simulated individual resulting in a given probability of contracting, dying from or surviving a range of conditions such as cancer, diabetes or heart disease.
• Events “compete” to occur in each simulated life and a random component embedded in the models ensures that not all individuals at risk of an event may experience it. Individual life trajectories are simulated until death.
• Costs can be assigned to interventions associated with the life events that have been simulated to project a future trend in health spending.

There are many advantages to the microsimulation-based approach to modelling, helping to inform decision makers to make the best decisions possible, which include:

• Allowing the testing of the potential impacts of policies and interventions, through “what if” scenarios and the related costs of implementation e.g. what is the expected impact on CVD incidence following the implementation of a tobacco duty escalator?
• Time, cost and ethical advantages of using simulation models vs real world experiments, which are not possible with population level policy interventions
• Enabling a wide set of comparisons to identify the most promising combinations of prevention (including policy interventions), screening and treatment approaches for different types of patients
• Going beyond the follow-up periods of typical clinical trials so that long-term outcomes can be compared 20, 30, 50 years into the future
• Going beyond narrow definitions of study outcomes to outcomes in real-world settings where patients are complex and remain at risk of developing a wide range of disease conditions (i.e. accounting for comorbidities)

Why is microsimulation modelling the right approach for NCD modelling?

According to the OECD “Only microsimulation allow testing “what if” scenarios of the impact of changes in lifestyles³”

Microsimulation approaches enable interventions that modify a population’s exposure to risk factors, such as tobacco smoking, unhealthy eating, alcohol consumption and lack of physical activity, to be quantified in terms of their impact on health expenditures of investments in public health programmes, clinical interventions or medical treatments.

In particular:

• When the interventions are targeted to specific subgroups, for example, to type 2 diabetics only or towards a specific low-income group
• When we need to know how a proposed intervention impacts on behaviours of different types of individuals. Individual behaviours are also modelled more easily at the level of individuals than for groups e.g. individual age, sex, disease state.

Critically, dynamic microsimulation is the only modelling approach that is applicable if an individual’s history matters. For example, an individual’s history of risk-taking behaviour, such as smoking, alcohol use and nutrition matters for the development of certain diseases. An individual’s history of disease matters for whether they live or die. Microsimulation models are designed to remember an individual‘s history and take it into account to influence their future life course.

The HealthLumen microsimulation model

HealthLumen’s microsimulation model has a long and well-established heritage.

The model was first developed for the UK Government’s Foresight Programme on “Tackling Obesities: Future Choices“ in 2005. The model has since been developed for over 70 national and regional populations covering over 20 NCDs including obesity, type 2 diabetes, coronary heart disease, COPD, stroke, hypertension, cancer, liver disease, chronic kidney disease, asthma, and dementia.

Extensively published and validated it is a leading model for determining trends in NCDs and risk factors, enabling the planning of interventions from prevention, management, treatment and long term care.

We model populations of up to 100 million people into the future, and examine how different kinds of health risks will affect them.

Our model works dynamically, tracking real change, showing how your chosen population’s health will evolve and transform over time.

References:

1. The Financial Burden of NCDs
2. The Economic Burden of Elevated Blood Glucose Levels in 2017: Diagnosed and Undiagnosed Diabetes, Gestational Diabetes Mellitus, and Prediabetes
3. A Comparative Analysis of Health Forecasting Methods