Abstract
BACKGROUND: Hospitals are central to healthcare systems and understanding the drivers of hospital activity is critical for effective capacity planning, especially amid demographic shifts and fiscal pressure. In the English National Health Service (NHS), there are plans to construct 40 new hospitals. There is limited evidence on factors driving hospital activity. This study provides retrospective estimates of the effects of population growth (P), changes in the age-sex structure (S), shifts in age-specific health status (H), and residual factors (R) on hospital activity in England. METHODS: Total annual growth is modelled as G = (1 + P)(1 + S)(1 + H)(1 + R) - 1. Negative binomial regression models using hospital episode statistics from 2011 to 2019 were used to decompose total annual growth rates (G) into P,S,H and R by five points of delivery (listed below). RESULTS: The annual growth rate for elective admissions was 2.29% (95% CI: 1.06% to 3.54%) which was made up of p = 0.88% (-0.96% to 2.67%), S = 0.69% (-0.62% to 2.18%), H = -0.03% (-0.23% to 0.17%) and R = 0.73% (0.58% to 0.88%). The annual growth rate for non-elective admission was 2.69% (1.55% to 3.83%) which was made up of p = 0.59% (-1.08% to 2.50%), S = 0.25% (-1.17% to 1.61%), H = 0.18% (-0.43% to 0.81%) and R = 1.65% (1.13% to 2.17%). The annual growth rate for maternity admissions was -0.70% (-5.65% to 4.51%) which was made up of p = 0.41% (-6.25% to 8.36%), S = -0.08% (-4.93% to 5.23%), H = 0.00% (NA) and R = -1.02% (-1.61% to -0.43%). The annual growth rate for outpatient attendance was 4.51% (3.62% to 5.40%) which was made up of p = 0.61% (-0.71% to 1.96%), S = 0.46% (-0.63% to 1.56%), H = -0.02% (-0.43% to 0.38%) and R = 3.41% (3.11% to 3.71%). The annual growth rate for emergency department attendance was 1.29% (0.71% to 1.88%) which was made up of p = 0.27% (-0.56% to 1.09%), S = 0.09% (-0.55% to 0.70%), H = 0.09% (-0.21% to 0.39%) and R = 0.84% (0.58% to 1.10%). CONCLUSIONS: In general, growth in hospital activity was mostly driven by residual (R) factors, followed by population growth (P). Age-specific health status (H) had the lowest impact. These data may provide useful context for planning future hospital activity in England.