A brief, global history of microsimulation models in health: past applications, lessons learned and future directions

This review discusses the evolution of microsimulation models in health over the past three decades. We focus on three aspects of health microsimulation. First, we describe the origins and applications of health microsimulation, including how early research challenged early methodologies and led to the development of more rigorous models. We describe limitations of early means-based models and how more detailed methods, which are based on health-specific input data, overcame many of the early data shortfalls and assumptions. Second, we discuss the global expansion of applications of microsimulation to health over the last ten years. Many health microsimulation models have focussed on health expenditure, the ageing population, diabetes, mortality modelling, and spatial models. Health models over the past few decades have expanded to many countries including Canada, Africa, the United States, the UK, Sweden, the Netherlands, New Zealand and Australia. Finally, we describe future developments, including emerging research fields for microsimulation and health and how the early development of health microsimulation models provides important lessons for emerging applications. These include the emerging field of genomics and precision medicine, and the diagnosis and treatment of childhood cancers and rare diseases. We suggest research directions, including the need for good data to avoid model errors, and highlight some pitfalls to avoid.