Forecasting a better health service

Medical Matters: 'Whoever wishes to investigate medicine properly, should proceed thus: in the first place consider the season of the year and what effects each of them produces, for they are not all alike," Hippocrates.

Whether it is an increase in the number of bone fractures after a period of freezing weather or a mini-epidemic of sunburn during a hot spell, the weather obviously affects our health.

Last year, researchers who looked at hospital admissions among women aged under 50 in 17 countries found a link between lower temperatures and a higher risk of cardiovascular diseases. A five degree drop in temperature was linked with a 7 per cent rise in admissions for stroke and a 12 per cent rise in admissions for heart attack, the British researchers concluded, speculating that changes in clotting mechanisms and blood pressure could be the reason.

At the other extreme of temperature, Taiwanese researchers reported in 2003 that high levels of air pollution, coupled with warm weather, increased the number of strokes by 50 per cent. They studied men and women living in the heavily industrialised city of Kaohsiung. On days when the temperature was greater than 20 degrees centigrade and pollution levels were high, the number of stroke admissions rose.

Nitrogen dioxide and particulate matter were the most significant pollutants associated with the increase, with the authors speculating that increased plasma viscosity and increased heart rate were to blame.

This link might explain some of the high mortality observed among older people in Paris in the summer of 2003 during a prolonged and intense heatwave. Increases in deaths have occurred in other cities, notably where summer climates are highly variable, so that residents have never been able to adapt to extreme heat.

Probably the most dangerous weather conditions occur when a mild winter spell is followed by a sharp drop in temperature. Research has shown that about two days after the onset of cold weather there is a sudden increase in the number of heart attacks. Some five days later, the number of people admitted to hospital with a stroke rises, while about 10-12 days after the temperature drops the number of patients with respiratory illnesses goes up.

With such well established patterns, it is perhaps not surprising that the Met office in Britain has developed a programme, Forecasting the Nation's Health, to see if it can predict surges in demand for healthcare and ultimately help the national health service with manpower planning.

According to Dr William Bird, a medical adviser to the Met office, "throughout England and Wales there is a 2 per cent increase in mortality for every degree the temperature drops below 18 degrees centigrade and snowfall doubles the number of heart attacks seen in hospital A&E departments".

Working with the Department of Health in London, Dr Bird and his colleagues combine information on respiratory infections, hospital admissions, GP workload and weather data to predict future demands on the health service, such as the likely number of emergency admissions by age group and the admission rates for different conditions in different regions.

"Forecasts of health admissions are issued twice weekly and enable bed managers to drive discharges, change staffing levels or cancel elective admissions with advance notice," according to the Met office.

Does it work? The Met office says that in the first years of operation the health forecast had an accuracy of 65 per cent overall, but that all the significant rises in workload in southern and northern areas of the UK were predicted.

While there are no plans to introduce a similar system here, it is probably something the HSE and Met Éireann should consider. Health forecasting will not solve our A&E crisis, but it may help even out some of the more severe clashes between elective and emergency demand.

But a system like this will never be 100 per cent accurate. As Brendan McWilliams, our Weather Eye, said last week, "there remains several reasons why weather forecasting can never be an exact science". He was responding to claims made by a Donegal businessman that an inaccurate forecast of bad weather had adversely affected hotel bookings.

I would much rather a system that erred on the side of caution. Whether for health service planning reasons or straightforward weather forecasting, it is better to accept a low level of false alarms rather than put people's lives at risk.

Dr Muiris Houston is pleased to hear from readers at mhouston@irish-times.ie but regrets he cannot answer individual queries.