Winter pressure is the most predictable crisis in the health calendar, yet it is planned for reactively almost every year. This tool produces a forward-looking Winter Vulnerability Index across five components, pairing NICE-evidenced interventions with a modelled estimate of when each is best deployed.
Cold weather converts manageable frailty into admission through a well-understood chain: cold homes, respiratory illness, falls on ice, and the seasonal collapse of informal care networks over the holidays. Because the mechanism is predictable and the interventions are evidenced, winter is the clearest case where acting on a forecast beats reacting to an outcome.
The index scores each district across frailty load, seasonal amplifiers, prescribing signals, system headroom and social isolation. Crucially it is forward-looking, built to inform deployment before the season rather than to describe it afterward. Each high-risk district carries a suggested intervention window: the interventions themselves are NICE-evidenced, while the timing of the window is a modelling inference from the seasonal pattern, not an empirically established optimum. It is a planning prompt, not a clinical instruction.
Forecasts inherit weather uncertainty. The index models seasonal risk structurally; it cannot predict the severity of a given winter, and a mild season will make it look over-cautious in hindsight.
Component weighting is a planning judgement. The relative weight of, say, cold-home risk versus system headroom is a defensible prior, not an empirical constant.
System headroom is the hardest component to measure openly. Capacity data lags and is incomplete, so this component carries the most uncertainty.