The number of admissions to non-cardiac ICU in those over 80 has risen by 63%, and if I were a journalist I’d stop reading there, but I’m not. The over 80s still make up the same proportion of all non-cardiac ICU admissions (13.8%) because the total number of ICU admissions has increased by 60% - surely when we talk about societal changes that is the headline.
We know that most healthcare usage is (perhaps obviously) in the last few months of life, with a suggestion that older people have a lower use of secondary care compared to people dying when they are younger. Perhaps therefore, investigating the elderly is not the right question. Age is a risk factor for death I accept, but is what we are seeing now just a consequence of there being a ‘peak’ in older people at present (baby-boomers etc.)? Experience (and the data) tells us that outcomes for elderly people in ICU are not as horrendous as we might be biased to presume, and if we can ‘do good’ then we will all continue to admit. What I would question is any societal changes towards accepting critical care admission as part of the dying process and a reluctance to accept medical futility; so a study similar to this looking at the impact of non-survivors rather than the elderly may make more sense. For an excellent review of the elderly in intensive care click here.
The number of ICU days taken by the elderly increased by a smaller proportion than the number of admissions, suggesting mean LOS to be getting shorter. Given that the proportion of ICU bed days taken by the elderly has not changed, this reduction in LOS would I suspect be seen across the board, in the young and old. This may well be an oversimplification however, as median LOS has not changed suggesting that there are still elderly patients with long stays (I've spent ages getting muddled about the median remaining the same and the mean reducing - it's a shame the authors didn't include frequency charts for LOS which would give a better representation than just the median and mean for a distribution that must be at least binomial)
Interestingly the elderly admitted to ICU don’t appear to be getting sicker (or better), so the threshold for admission maybe hasn’t changed too much.
Just to finish on a bit of pedantry, some of the data is population data and not from a sample (e.g. the percentage of the Dutch population over 80yrs of age). The authors quote confidence intervals, which doesn’t make sense if the population statistic is known (a confidence interval is a range within which the population statistic is probable to lie).
By way of suggestions for comment:
- If we could accurately assess frailty is age irrelevant?
- Would evidence based critical care for the over 80s be different to what we currently practice?
- Is the Netherlands so different to the UK that the results are not of relevance to our practice?
- What other data about the over 80s would you like to see?
As always I look forward to reading your thoughts…