Prevalence of ischaemic heart disease at admission to intensive care and its influence on red cell transfusion thresholds: multicentre Scottish Study

Abstract

Restrictive transfusion triggers are safe for most critically ill patients, but doubts exist for patients with ischaemic heart disease (IHD). We investigated the prevalence of reported IHD at admission to the intensive care unit (ICU) and investigated how this influenced red cell transfusion triggers. We also compared observed practice with the clinicians' responses to clinical scenarios.We studied 1023 sequential ICU admissions over 100 days to 10 Scottish ICUs. Daily haemoglobin, red cell transfusion, and haemorrhage data were available for 99.4% of 5638 ICU patient days. We recorded if IHD was recorded in clinical records at ICU admission. We grouped admissions as having a non-cardiac primary ICU diagnosis and no documentary evidence of IHD (Group 1, n=697), a non-cardiac primary ICU diagnosis with evidence of IHD (Group 2, n=213), or a cardiac primary ICU admission diagnosis (Group 3, n=113). We examined pre-transfusion haemoglobin concentration (Hb) for transfusion episodes not associated with haemorrhage. Clinical transfusion scenarios were sent to intensivists in the ICUs after data collection, which were designed to explore the clinicians' attitude to transfusion triggers in patients with IHD.Previous myocardial infarction was documented in 159 (16%), cardiac failure in 142 (14%), and angina in 167 (16%). Overall, 28.8% of admissions had >/=1 of these documented. The adjusted mean (se) pre-transfusion Hb concentrations varied across the groups. These were 74 (2.2) g litre(-1) in Group 1, 77 (2.3) g litre(-1) in Group 2, and 79 (3.1) g litre(-1) in Group 3 (P=0.003 across the groups). There was concordance between observed practice and responses to the scenario similar to Group 1, but discordance for patients with IHD (Groups 2 and 3). In scenario responses, intensivists stated these patients should have significantly higher transfusion triggers than were actually observed (median [IQR] response for both groups: 90 [80-100] g litre(-1)).About 29% of patients admitted to Scottish ICUs had documented IHD, which was associated with small adjustments to Hb transfusion triggers. In response to scenarios, clinicians believe that patients with IHD require higher transfusion triggers than are observed in practice.