Epidemiology and outcomes of older patients admitted to Scottish intensive care units: a national database linkage study.

BACKGROUND

As the general population ages and life expectancy increases, health-care use by elderly people increases, including intensive care. Rationing and variation of access are ethically and politically challenging. We aimed to characterise the population-based incidence of intensive care unit (ICU) admissions of elderly people in Scotland; compare ICU admission and mortality between elderly and younger populations; and compare treatment intensity between these groups.

METHODS

We extracted complete, national 6-year cohort Scottish ICU admissions (Jan 1, 2005, to Dec 31, 2010) from the Scottish Intensive Care Society Audit Group database, which we linked to hospital Scottish Morbidity Record (SMR01) and death records. Annual incidence of ICU admissions of people aged 80 years or older was standardised for sex and socioeconomic status to the standard Scottish population (≥80 years) 2005-10. We compared mortality of elderly and younger people (<65 years) using the log-rank test.

FINDINGS

During 2005-10, 47 779 people were admitted to ICU (4561 patients ≥80 years [9·5%, 35·0/10 000 population], 26 784 patients <65 years [56·1%, 13·2/10 000]). Incidence of ICU admissions of elderly people fell from 36·6/10 000 population (95%CI 34·0-39·2) in 2005 to 30·3/10 000 (28·0-32·5) in 2010. ICU mortality was higher in elderly than in younger people (26·4% vs 16·1%, p<0·0001) as was 6-year mortality (68·0% vs 34·5%, p<0·0001). 2110 (80%) of 2627 elderly survivors were discharged home (younger 92%, 19 221/20 902), with a further 373 (14·2%) given rehabilitation (younger 1063, 5·1%) (χ(2)=525, p<0·0001). Age was an independent predictor of mortality (odds ratio 1·46, 95% CI 1·23-1·73, p<0·0001) after adjustment for confounders. In the pneumonia subgroup (elderly 294, younger 2167), mean acute physiology scores were similar (17·0 [SD 6·4] vs 17·6 [6·6]), organ support was higher in the elderly patients (77·0% vs 68·1%, p<0·0001), and median ICU length of stay was lower (6 days [IQR 3-13] vs 8 [3-16], p<0·0001).

INTERPRETATION

This study has shown that, by contrast with previously published research, admission rates of elderly people in Scotland fell between 2005 and 2010. Only the fittest elderly individuals were admitted to ICU, where initially they received a higher intensity of treatment than did younger patients; however, duration of ICU stay was shorter. Mortality rates were high, and age was an independent predictor of mortality.

FUNDING

Funding assistance for AD's MPH from Scottish Intensive Care Society, Scottish Society of Anaesthetists, Edinburgh Anaesthetics Research and Education Fund.