In trying to make sense of this this complex difficult and uncomfortable area however, one of the first steps must be to quantify and describe the risk.
Unplanned admissions to critical care would seem a reasonable place to look; sampling such cases for review is an accepted method that in an ideal world all critical care units would do in co-operation with the rest of the hospital.
830 records (unplanned admissions) were reviewed. In 56% of these an adverse event had occurred (to meet inclusion these events were judged to have a causal relationship to the admission). Perhaps surprisingly, in 98% of the 56%, only one adverse event had occurred (not entirely ‘Swiss cheesy’)
The patients admitted after an adverse event were not a well group. 44% were ASA 4 (why they used ASA I don’t know), 66% had functional limitation, and 46% were on their second hospital admission in 3 months. 71% of the patients were prescribed more than five chronic medications. 98% of those who went on to die did not have a prior DNACPR order in place.
The incidence of an unplanned admission involving an adverse event was approx. 120/100 000 patient days. This figure is not particularly meaningful, but as a rough guide if you could extrapolate (which you can't):
SRH has about 1000 beds. If all were full at all times, that’s 365 000 bed days/yr., giving 438 unplanned admissions to the ICCU p.a., about half of all our admissions.
The paper goes on to report that 46% are highly preventable, so if we could magically stop making avoidable errors, that would be about four admissions a week avoided (accepting that this is based on some dubious assumptions and approximations).
The cause groups were interesting, with 25% of adverse events being medication related (the majority apparently due to antibiotics and antithrombotics – I don’t have any details but I’d guess at the wrong antibiotic, or it being given too late, and PE’s). Apparently 100% of medication errors were deemed preventable which I suppose is true, but I think is an over-simplification.
24% were complications of surgery - with 27% preventable. I’m not a surgeon but that sounds a bit harsh??
12% of adverse errors were grouped as delays in or incorrect diagnosis – I think this helps highlight the case that adverse events are complex – we often have the joy of hindsight by the time the patient reaches critical care after a ward stay.
System issues were only causal in 12%, I think we anecdotally attribute this more often (not least because they’re a nameless person’s fault!).
19% of the adverse events were deemed to cause harm lasting greater than 1 year, but the attributable mortality was not increased. It was also interesting to note that 40% of the survivors were readmitted within 6 months (and 20% within 1 month - for the significance of 1 month click here) – whether that says more about background co-morbidity or the consequences of the adverse event I don’t know.
Please feel free to leave a comment below for discussion, but for further reading and resources: http://chfg.org/