Abstract T P138: Time Trends over Five Decades in Rates of Hospital Admission for Any Stroke and Stroke Subtypes in England

Abstract

Aim: To study 43 year trends in England in the rates of hospital admission for any stroke, and for stroke subtypes divided into stroke that was coded as haemorrhagic, ischaemic or of undefined etiology. Methods: We analysed routine hospital statistics from the Hospital In-Patient Enquiry and Hospital Episode Statistics to estimate trends in hospital admissions for any stroke and the three stroke subtypes in England from 1968 to 2011. The rates were age-standardised and expressed per 100 000 population. Results: The study included 2 555 180 admissions for stroke. Annual admission rates for any stroke increased from 169.79 (95% CI 165.61-173.97) in 1968 to 272.4 (270.86-273.93) in 1993; and then started to decline to reach 196.58 (195.38-197.79) in 2011. Admission rates for haemorrhagic stroke declined from 30.40 (28.61-32.19) in 1968 to 15.07 (13.80-16. 34) in 1982, and then steadily increased to 44.65 (44.07-45.24) in 2011. Admission rates for ischaemic stroke halved between 1968 and 1982 from 60.3 (57.81-62.80) to 32.32 (30.53-34.11). This was followed by a steep rise to 128.10 (127.13-129.07) in 2011. Admission rates for stroke of undefined cause increased from 75.54 (72.75-78.33) in 1968 to 178.35 (174.35-182.34) in 1984. This was followed by a sharp decline to 27.66 (27.22-28.11) in 2011. Conclusions: Considering any stroke in England, in the 1970s and1980s there was an increase in admission rates, followed by a decline in the 1990s, and a plateau in the first decade of the 21st century. Changes in stroke rates for haemorrhagic and particularly for ischaemic stroke tended to go in opposite directions from rates in stroke of unspecified type. Specifically, the observed declines in haemorrhagic and ischaemic stroke rates between 1968-1983 coincided with a rise in stroke of undefined cause. These trends reflect shifts in diagnostic preferences towards the latter at that time. The recent dramatic increase in ischaemic stroke rates is mirrored by a sharp fall in stroke of unspecified type. These changes are likely to result from the increasingly wide use of brain imaging. Trends in admission rates for a particular subtype of stroke should not be considered on its own, as changes in diagnostics, and preferences for diagnostic terminology, are likely to affect the rates.