Interpreting hospital admissions data across the Körner divide: the example of diabetes in the north western region.

Hospital admission rates and hospital bed utilization rates are presented for all admissions assigned to diabetes mellitus as principal cause in the North Western Region from 1980/81 to 1990/91. Data are derived from Hospital Activity Analysis (1980/81-1986/87) and from the Regional Information System (1988/89-1990/91). Admission rates for all categories of diabetes combined showed little change until 1986/87 after which they rose progressively each year. Hospital bed utilization rates fell progressively from 1986/87 onwards, the number of bed days attributed to diabetes falling from 1.9% to 0.8% of the total for all causes. Most of the rise in admission rates from 1988/89 onwards was attributable to diabetes with ophthalmic complications (ICD 250.4). This was heavily influenced by the enumeration of day cases in the Regional Information System. When these are discounted, admission rates for the more recent years are stable. Diabetes without mention of complications (ICD 250.0) showed the most dramatic decline in its contribution to hospital bed utilization. This may be the result of increasing outpatient support from diabetes specialist nurses whose numbers increased in inverse proportion to the fall in bed utilization attributed to ICD 250.0.