Back

Projected population health, demographics, and associated healthcare resource requirements from 2025 to 2047: modelling study in England

Shaw, L. F.; Onen-Dumlu, Z.; Vasilakis, C.; Wood, R. M.

2025-12-23 health economics
10.64898/2025.12.22.25342817 medRxiv
Show abstract

ObjectivesThere is a deficit of information available to guide longer-term healthcare planning in meeting health needs and ensuring financial sustainability. This study attempts to address this gap through projecting future population health state and demographics and associated healthcare resources required to satisfy expected demands over the next two decades. MethodsA mathematical model is developed for projecting a populations future age and health state, subject to births, deaths, immigration and emigration. These modelled outputs are combined with healthcare resource utilisation profiles to provide future cost estimates at a total level and for constituent healthcare settings. The model is calibrated to linked longitudinal patient-level data from a one-million-resident NHS healthcare system in England covering a mixture of urban, rural and coastal locations. ResultsCost growth is projected to exceed population growth over the period from 2025 to 2047, with costs growing by 15-22%, depending on the scale of international migration, and the population growing by 4-16%. This is being driven by a projected 15-30% growth in the size of all but the healthiest population health states. If budgets were constrained to population growth, then either a 4% productivity improvement would be required or a 10% reduction in those transitioning to worse health states. ConclusionsAn ageing and unhealthier population is leading to a 5-11% rise in average per-person healthcare costs. To avert this, healthcare administrators may need to prioritise the quicker-acting cash-releasing productivity schemes to fund the prevention measures that may ultimately be required to secure longer-term financial sustainability.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.