Limitations and practicalities of CAM-ICU implementation, a delirium scoring system, in a Dutch intensive care unit

BACKGROUND: Delirium is a frequently missed diagnosis in the intensive care unit (ICU). Implementation of the Confusion Assessment Method for the ICU (CAM-ICU) may improve recognition of delirium. However, the ICU team may be reluctant to adopt daily assessment by a screening tool. This report focusses on the obstacles and barriers encountered with respect to organisational context and prevailing opinions and attitudes when implementing the CAM-ICU in daily practice in a Dutch ICU. METHODS: A structured implementation process was set up comprising four phases: (1) assessing the current situation to understand behaviour towards delirium; (2) the identification of barriers to the implementation of the CAM-ICU; (3) preparation of the ICU team for a change in attitude; and (4) evaluation of the effects of implementation. RESULTS: Phase 1 demonstrated that there was no delirium protocol available; it was left to the attending physicians when and how to diagnose delirium in each individual patient. In addition, nurses acted on delirium in a non-structured way; nurses thought implementation of the CAM-ICU would be very time-consuming and would not add to their ability in recognising delirium. In Phase 2, several barriers to implementation were addressed. Firstly, all nurses had to be convinced that delirium is an important problem and, secondly, logistics had to be put in place, for example, picture cards at every bedside, communication between daily nurses and a delirium working group had to be improved. In Phase 3, 10 nurses were educated to perform the CAM-ICU through several training sessions which included videos to illustrate different delirium states; these trained nurses educated all other nurses. A check box in the daily records was introduced to denote whether the CAM-ICU had been performed. In Phase 4, after a training period and 2 months of actual routine bedside CAM-ICU performance, evaluation demonstrated that frequency of assessments on un-sedated patients had increased from 38% to 95% per nursing shift. A short survey amongst the ICU nurses also showed that awareness of delirium and appreciation of the clinical problem had markedly increased. CONCLUSION: Implementation of the CAM-ICU in daily critical care is feasible. A structural training programme is probably helpful for success of implementatio