Demographics of critically ill transferred patients

Abstract

There are few reports describing the demographic details and outcome of critically ill patients transferred into intensive care units (ICU) from other hospitals. We analysed the SICSAG database containing data which are collected on consecutive admissions to adult general ICU in Scotland. Between January 1, 1995 and December 31, 1997 there were a total of 20 738 ICU admissions, of which 1805 (8.7%) were admitted after transfer. A total of 106 patients were excluded from the severity scoring. Transferred patients had a median age of 55 yr with a range of 16–89 yr. The mean Acute Physiology And Chronic Health Evaluation (APACHE) II score for the first 24 h of ICU admission was 18.3 and the Standardized Mortality Ratio (SMR) was 0.94 with 95% confidence intervals (CI) of 0.88–1.0. For hospital survivors the mean APACHE II score was 16. A total of 509 (30%) patients died in hospital and in this group the mean APACHE II score was 23.7. The median length of stay (LOS) in ICU was 4.2 days, and an interquartile range (IQR) of 1.55–10.0 days in all patients. In survivors the median LOS in ICU was 4.6 days, with an IQR of 1.7–10.2 days, compared with a median LOS of 3.1 days (IQR of 1.1–9.6) in ICU non-survivors. There were 105 different diagnostic categories for transferred patients. These have been further classified into: respiratory 24.5%; cardiovascular 21.2%; trauma 14.8%; gastrointestinal 12.1%; neurological 9.7%; other 6.4%; sepsis 5.5%; renal 3.6%; hepatic 1.4% and haematological 0.7%. It is surprising that only 3.6% of transferred patients had a primary renal disorder. This may reflect that the increasing use of haemofiltration in ICU has led to a reduction in tertiary referral in these patients. This preliminary analysis also demonstrates that 45.7% had a primary respiratory or cardiovascular diagnosis, compared to 28% reported previously.1 However, the reason for admission to ICU is not necessarily the same as the reason for transfer. Further work is required to identify relevant differences in demographics and outcome between this group and patients who have not been transferred.