Should octogenarians and people aged 90 years and over be treated separately in studies of stroke? An evaluation of national audit data.

Introduction: Population studies frequently use ≥80 years for defining 'very old' but as mean life expectancy frequently exceeds 80 years internationally, this may no longer be appropriate. Those ≥90 years now represent a significant proportion of stroke patients. We examined national data to examine the differences between those 80-89 years and those 90+ years.Methods Data from the Irish National Audit of Stroke (2017-22 inclusive), including demographic, admission and outcome data, including prestroke and discharge modified Rankin Scores (mRS), were analysed. Proportional data were analysed using Chi-square statistics.

Results: Data on 26 829 individual stroke events were analysed of which 7329 (27.3%) were in people 80-89 years; 52.8% were women. 1708 events occurred in people ≥90 (6.4%); 70.0% were women. 73.7% of those 80-89 years had mRS < 3 prestroke vs 51.3% of those ≥90 (P < .001). In hospital mortality for people ≥90 was higher (26.8% vs 17.4% P < .001) and they were less likely to have mRS < 3 at discharge (17.0% vs 35.8% P < .001). Proportion of haemorrhagic stroke was significantly lower in those ≥90 (15.3% vs 12.9% P = .015). Only one haemorrhage was reported amongst 31 people ≥100 years. The proportion of atrial fibrillation (AF) detected following stroke was not significantly different (≥90 years: 33.9%, 80-89 years, 32.4% P = .38).On logistic regression, nonrecovery to independence (mRS >2) in those ≥90 was associated with prestroke mRS, haemorrhagic stroke, AF and being thrombolysed.

Conclusion: There are differences in profiles and outcomes between the groups, and it is now more appropriate to consider them separately.